System and method for generating machine learning models for asset planning

Abstract

An approach for utilizing machine learning models for asset planning is disclosed. The approach includes receiving historical performance data and economic data associated with one or more illiquid assets. The approach also includes training a machine learning model using the historical performance data to predict one or more features including cumulative contributions, distributions, and net asset value (NAV) percentage changes. Further, the approach includes incorporating regime-based transition probabilities into the machine learning model to account for different economic stages; and generating cash flow forecasts and stress testing output for the illiquid assets based on the trained model.

Description

Attorney Docket No.: P10355PC00 PatentSYSTEM AND METHOD FOR GENERATINGMACHINE LEARNING MODELS FORASSET PLANNINGRELATED APPLICATIONS

[0001] This application claims the benefit of U.S. Provisional Patent Application No.63 / 556,251, titled “System and Method for Generating Machine Learning Models for AssetPlanning,” filed February 21, 2024, the entire disclosure of which is hereby incorporated byreference herein.BACKGROUND

[0002] Despite the current set of tools, data analysts in all industries face the daunting task ofanalyzing an overwhelming barrage of information to determine meaningful, impactful data. Oneparticular sector involves the financial industry, such as wealth management. Both wealthmanagers and associated consumers share the responsibility for achieving monetary targets. Theselection of vast number of profitable assets and investment portfolios require allocation ofenormous resources in personnel and technology. Existing systems that wealth managers rely onfor financial analysis have not kept based with development in data processing technology. Thechallenge is magnified when the data (e.g., news, social media, etc.) significantly affects marketconditions in real-time. Hence, accurate evaluation of assets or portfolios remains a difficultproblem even with the recent advancement of data analytics and artificial intelligence. Aparticularly challenging area concerns illiquid assets, which are difficult to convert to cash withoutsignificant losses, and are difficult to value due to large price swings.SOME EXAMPLE EMBODIMENTS

[0003] Therefore, there is a need for an approach that enables readily assessing performanceof assets or portfolios involving illiquid assets by leveraging machine learning.Attorney Docket No.: P10355PC00 Patent

[0004] According to one embodiment, a method comprises receiving historical performancedata and economic data associated with one or more illiquid assets. The method also comprisestraining a machine learning model using the historical performance data to predict one or morefeatures including cumulative contributions, distributions, and net asset value (NAV) percentagechanges. Further, the method comprises incorporating regime-based transition probabilities intothe machine learning model to account for different economic stages; and generating cash flowforecasts and stress testing output for the illiquid assets based on the trained model..

[0005] According to another embodiment, an apparatus comprises at least one processor, andat least one memory including computer program code for one or more computer programs, the atleast one memory and the computer program code configured to, with the at least one processor,cause, at least in part, the apparatus to receive historical performance data and economic dataassociated with one or more illiquid assets. The apparatus is also caused to train a machine learningmodel using the historical performance data to predict one or more features including cumulativecontributions, distributions, and net asset value (NAV) percentage changes. The apparatus is alsocaused to incorporate regime-based transition probabilities into the machine learning model toaccount for different economic stages. The apparatus is also caused to generate cash flow forecastsand stress testing output for the illiquid assets based on the trained model.

[0006] In addition, for various example embodiments of the invention, the following isapplicable: a method comprising facilitating a processing of and / or processing (1) data and / or (2)information and / or (3) at least one signal, the (1) data and / or (2) information and / or (3) at least onesignal based, at least in part, on (or derived at least in part from) any one or any combination ofmethods (or processes) disclosed in this application as relevant to any embodiment of theinvention.

[0007] For various example embodiments of the invention, the following is also applicable: amethod comprising facilitating access to at least one interface configured to allow access to at leastone service, the at least one service configured to perform any one or any combination of networkor service provider methods (or processes) disclosed in this application.

[0008] For various example embodiments of the invention, the following is also applicable: amethod comprising facilitating creating and / or facilitating modifying (1) at least one device userAttorney Docket No.: P10355PC00 Patentinterface element and / or (2) at least one device user interface functionality, the (1) at least onedevice user interface element and / or (2) at least one device user interface functionality based, atleast in part, on data and / or information resulting from one or any combination of methods orprocesses disclosed in this application as relevant to any embodiment of the invention, and / or atleast one signal resulting from one or any combination of methods (or processes) disclosed in thisapplication as relevant to any embodiment of the invention.

[0009] In various example embodiments, the methods (or processes) can be accomplished onthe service provider side or on the mobile device side or in any shared way between the serviceprovider and mobile device with actions being performed on both sides.

[0010] For various example embodiments, the following is applicable: an apparatuscomprising means for performing a method of any of the claims.

[0011] Still other aspects, features, and advantages of the invention are readily apparent fromthe following detailed description, simply by illustrating a number of particular embodiments andimplementations, including the best mode contemplated for carrying out the invention. Theinvention is also capable of other and different embodiments, and its several details can bemodified in various obvious respects, all without departing from the spirit and scope of theinvention. Accordingly, the drawings and description are to be regarded as illustrative in nature,and not as restrictive.BRIEF DESCRIPTION OF THE DRAWINGS

[0012] The embodiments of the invention are illustrated by way of example, and not by wayof limitation, in the figures of the accompanying drawings:

[0013] FIG. 1 is a diagram of an AI-based decision support platform, according to oneembodiment;

[0014] FIG. 2 is a diagram of the components of the AI-based decision support platform ofFIG. 1, according to one embodiment;

[0015] FIG. 3 is a diagram of the AI-based decision support platform of FIG. 1 interactingwith various data sources, according to one embodiment;Attorney Docket No.: P10355PC00 Patent

[0016] FIG. 4A is a flowchart of a process for generating a unifying index by the AI-baseddecision support platform of FIG. 1, according to one embodiment;

[0017] FIG. 4B is a flowchart of a process for generating a relative measure indicator by theAI-based decision support platform of FIG. 1, according to one embodiment;

[0018] FIGs. 4C and 4D are flowcharts of processes for utilizing machine learning models forasset planning by the AI-based decision support platform of FIG. 1, according to variousembodiments;

[0019] FIGs. 5A-5H are diagrams of a graphical user interface (GUI) relating to datamanagement processes performed by the AI-based decision support platform of FIG. 1, accordingto one embodiment;

[0020] FIGs.6A and 6B are diagrams of a GUI relating to asset allocation processes performedby the AI-based decision support platform of FIG. 1, according to one embodiment.

[0021] FIGs. 7A-7D are diagrams of a GUI relating to security selection processes and indexgeneration processes performed by the AI-based decision support platform of FIG. 1, according toone embodiment;

[0022] FIG. 8 is a diagram of a neural network that can be implemented by the AI-baseddecision support platform of FIG. 1, according to one embodiment;

[0023] FIG. 9 is a diagram of hardware that can be used to implement various exampleembodiments;

[0024] FIG. 10 is a diagram of a chip set that can be used to implement various exampleembodiments; and

[0025] FIG. 11 is a diagram of a mobile terminal (e.g., handset) that can be used to implementvarious example embodiments.Attorney Docket No.: P10355PC00 PatentDESCRIPTION OF SOME EMBODIMENTS

[0026] Examples of a method, apparatus, and computer program for training a machinelearning model to assess performance of illiquid assets or portfolios are disclosed. In the followingdescription, for the purposes of explanation, numerous specific details are set forth in order toprovide a thorough understanding of the embodiments of the invention. It is apparent, however,to one skilled in the art that the embodiments of the invention may be practiced without thesespecific details or with an equivalent arrangement. In other instances, well-known structures anddevices are shown in block diagram form in order to avoid unnecessarily obscuring theembodiments of the invention.

[0027] As noted, illiquid assets (e.g., private equity, real estate, infrastructure, etc.) poseadditional challenges when determining their performance. For example, the need for a companyto quickly raise funds to meet financial obligations is problematic with such assets. Also, pricevolatility is an issue for financial reporting, as well as lack of transparency as such assets are notpublicly traded. Moreover, illiquid assets are complex to value and manage.

[0028] FIG. 1 is a diagram of AI-based decision support platform, according to oneembodiment. To address the noted drawbacks of conventional systems and approaches toprocessing the vast amount of data for proper selection of investments, a system 100 of FIG. 1includes an AI-based decision support platform 101 that introduces the capability to assess and tomeasure real-time impact of an asset / security or portfolio. Specifically, with respect to assetplanning, the platform 101 utilizes two machine learning models (described below): (1) a two-stepprediction model that leverages the deterministic approach as an anchor with an error predictionapproach, and (2) a Cash Flow Prediction Model (CFPM) that applies machine learningtechniques.

[0029] Illiquid assets (i.e., private equity, real estate, infrastructure, etc.) have attractedtrillions of dollars in commitments to multi-year locked-in strategies (across single asset andcommingled fund structures). This commitment allure (driven by a myriad of factors includinghigher expected returns, longer management periods for thesis execution, and lower markedvolatility) continues to outpace the financial models needed to manage the assets for years afterAttorney Docket No.: P10355PC00 Patentthe legal commitment. Taking pause to this commit-and-forget approach, it is recognized that newfinancial models for illiquid asset cash flow forecasting and stress testing that should allow formore robust asset planning. Accordingly, the platform 101 provides: a) improved, better-fittedmodels can help better manage cash flow profiles (e.g., holding or releasing cash reserves, timinginvestments, etc.); b) model choices give optimality for fitting uniquenesses and nuances (e.g.,vintages, features, profiles, etc.); c) unified stress testing and simulations can help risk managecash flow profiles in unison with other asset classes; d) augmenting analysis with synthetic databridges the data gaps; and e) overall these elements also can initiate a framework for applying AIto other investment aspects such as sector signals, product or investment selection, warnings andallocations themselves.

[0030] Additionally, the platform 101 can supplement and / or integrate with a third partysystem 103, which may be a conventional financial analysis system that is utilized by financialadvisors. The platform 101 can parse through numerous datasets, models, viewpoints, visuals, etc.concurrently to continually assess historic and predicted performance within each aspect of theasset life cycle, subject to defined or suggested objective functions. By way of example, theplatform 101 can supplement the financial advisor’s experience-based, traditional portfoliodecisions with Artificial Intelligence-powered visualizations for assessing a Security or Portfolio(A), evaluating the predicted path profile of A, comparing A with B (and transition of A to B), andmeasuring real-time impact on A. With respect to the third party system 103, the platform 101integrates within the technology infrastructure of the system 103 to execute user-defined oroptimized objective functions and can also run the processes on existing portfolios.

[0031] Although the platform 101 is explained utilizing use cases involving applications to thefinancial sector, it is contemplated that the platform 101 can unify other forms of measures / metrics(e.g., operational performance, etc.).

[0032] Among other functions and features, the platform 101 can harness research, data,programming and advanced AI methodologies to address daunting development possibilitiescentered around the following: removing the drudgery of complex tasks to drive efficiency;Attorney Docket No.: P10355PC00 Patentproviding visuals that are interpretable to facilitate more creative work; and determining meaningresearch data.

[0033] Common measures, e.g., the International System of Units (SI), known as the metricsystem, are widely used international standard for measurement. A factor for the adaptation ofthis or any measure has to do with its simplicity and explainability. Thus, the approach describedherein accounts for such explainability to arrive at a unifying framework for performancemeasures.

[0034] Regarding performance measures, practitioners, researchers, etc. have proposedvarious measures for assessing assets and / or portfolios. These range from simplistic measures tothose requiring more advanced financial engineering techniques that propose higher variabilityand situational capture. Most can be classified as,^ Using historic or forecasted data for estimates (e.g., Sharpe, cVAR, AcceptabilityIndex, Max Draw-Down, etc.)^ Simulations with known or what-if scenarios (e.g., historic, beta-based, factor-based,PCA-based, etc.)^ Measured as a unit as in a ratio, percentage, number, etc.^ Assessed as a single path or from (or subset of) distribution(s)^ Having indeterminate or fixed holding period(s) or maturities^ Evaluating market, credit, liquidity, etc.^ Evaluating regimes, factor, macro, etc.^ Assessing holding data, alternative data, operational data, etc.

[0035] For most practitioners, the assessment of an asset or portfolio hinges on the relativerange of the individual evaluation measures of the assets or portfolios. This relative range whenlooked at as a visual grid with selected measures and weights facilitate the practitioners' preferencesin assessing the asset or portfolio. Where some form of back testing and / or simulation gives thepractitioner the comfort around the choice of measures, its weight and consequently the asset orAttorney Docket No.: P10355PC00 Patentportfolio being evaluated. As business extensions, practitioners have started to bring in prospect theory,behavioral economics, profiling, decision analysis, experiential, etc. to include risk capacity, toleranceand such for potentially better audience engagement; whereby proxies that try to knit together thebehavioral and portfolio aspects / relationships are also being proposed. While engagement isencouraging, it should be noted that proxies can have dilution risk and / or add noise. As a baseline, theproxy needs to capture the multi-dimensionality of the asset and / or portfolio and simplistic, uni-dimensional approaches have the risk of underlying biases and / or instilling unwarranted comfort whilenot being complete (e.g., only using VaR or single stress as a proxy can be misleading).

[0036] According to various embodiments, the platform 101 generates a composite index or aunifying index (Explainability Index (EI)) that helps evaluate asset or portfolio performance overa historic or forecasted time period. The Explainability Index explicitly balances different inputcategories of performance measures according to default or specified preferences and gives acomposite bounded score within a predefined or predetermined range, e.g., between 0 and 1. TheExplainability Index can incorporate all desired relative / absolute performance measures -- e.g.,measures that focus on volatility, drawdowns, returns, or other aspects. The Explainability Indexgives a unified view as well as the contribution of each performance measure or category.Additionally, the platform 101 can determine a relative measure indicator (i.e., relative measure(Risk of Target (RoT))) that leverages the Index for comparing the performance of assets / portfoliosamong themselves or against benchmarks (as targets). This composite and component analysisprovides perspectives on the drivers of divergence (as a point-in-time, trend, related or relativeassessment). Further details of the approach are described in FIGs. 2-4.

[0037] As shown in FIG. 1, the system 100 also comprises user equipment (UE) 105a-105n(collectively referred to as UE 105) that may include or be associated with applications 107a-107n(collectively referred to as applications 107). In one embodiment, the UE 105 has connectivity tothe AI-based decision support platform 101 via the communication network 109. Under certainscenarios, the AI-based decision support platform 101 performs one or more functions associatedwith data analysis in conjunction with the UEs 105a-105n. By way of example, the UE 105 is anytype of mobile terminal, fixed terminal, or portable terminal including a mobile handset, station,unit, device, multimedia computer, multimedia tablet, Internet node, communicator, desktopAttorney Docket No.: P10355PC00 Patentcomputer, laptop computer, notebook computer, netbook computer, tablet computer, personalcommunication system (PCS) device, personal navigation device, personal digital assistants(PDAs), audio / video player, digital camera / camcorder, positioning device, television receiver,radio broadcast receiver, electronic book device, game device, a smartphone, a smartwatch, smarteyewear, or any combination thereof, including the accessories and peripherals of these devices,or any combination thereof. It is also contemplated that the UE 105 can support any type ofinterface to the user (such as “wearable” circuitry, etc.). In one embodiment, the UE 105 mayinclude Global Positioning System (GPS) receivers to obtain geographic coordinates fromsatellites (not shown) for determining current location and time associated with the UE 105; suchGPS information can be utilized to geo-tag images captured by UE sensors (not shown). The UE105 is capable of supporting a graphical user interface (GUI) that provides the GUI of FIGs. 5-7.

[0038] The AI-based decision support platform 101 operates in conjunction with one or moreapplications resident on an UE 105. By way of example, the applications 107 may be any type ofapplication that is executable at UE 105, such as content provisioning services, camera / imagingapplication, media player applications, social networking applications, calendar applications, andthe like. In one embodiment, the applications 107 may assist in conveying sensor information viathe communication network 109. In another embodiment, one of the applications 107 at the UE105 may act as a client for the AI-based decision support platform 101 and perform one or morefunctions associated with the functions of the platform 113 by interacting with the platform 113over the communication network 109.

[0039] One or more data sources 109a-109n are accessible via the network 109 by the platform101. The data sources 109a-109n can include any content relevant to making financial decisionsrelating to securities, such as market data (e.g., historical or real-time market information), news,financial holdings, alternative data (as shown in FIG. 3). The retrieved data can reside withindatabase 111 of the AI-based decision support platform 101. It is contemplated that database 111can be implemented as a cloud storage system.

[0040] The communication network 109 of system 100 includes one or more networks suchas a data network, a wireless network, a telephony network, or any combination thereof. It isAttorney Docket No.: P10355PC00 Patentcontemplated that the data network may be any local area network (LAN), metropolitan areanetwork (MAN), wide area network (WAN), a public data network (e.g., the Internet), short-rangewireless network, or any other suitable packet-switched network, such as a commercially owned,proprietary packet-switched network, e.g., a proprietary cable or fiber-optic network, and the like,or any combination thereof. In addition, the wireless network may be, for example, a cellularnetwork and may employ various technologies including 5G (5th Generation), 4G, 3G, 2G, LongTerm Evolution (LTE), enhanced data rates for global evolution (EDGE), general packet radioservice (GPRS), global system for mobile communications (GSM), Internet protocol multimediasubsystem (IMS), universal mobile telecommunications system (UMTS), etc., as well as any othersuitable wireless medium, e.g., worldwide interoperability for microwave access (WiMAX), codedivision multiple access (CDMA), wideband code division multiple access (WCDMA), wirelessfidelity (Wi-Fi), wireless LAN (WLAN), Bluetooth®, Internet Protocol (IP) data casting, satellite,mobile ad-hoc network (MANET), and the like, or any combination thereof.

[0041] In one embodiment, the AI-based decision support platform 101 may be a platformwith multiple interconnected components. The AI-based decision support platform 101 mayinclude multiple servers, intelligent networking devices, computing devices, components andcorresponding software for providing real-time data analysis. In addition, it is noted that the AI-based decision support platform 101 may be integrated or separated from services platform. Also,certain functionalities of the system 101 may reside within the UE 105 (e.g., as part of theapplications 107).

[0042] Moreover, the platform 101 can interface with various services systems (not shown),such as notification services, content (e.g., audio, video, images, etc.) provisioning services,application services, storage services, contextual information determination services, socialnetworking services, location-based services, information-based services, etc.

[0043] By way of example, UE 105, the AI-based decision support platform 101, the thirdparty system 103 with each other and other components of the communication network 109 usingwell known, new or still developing protocols (e.g., IoT standards and protocols). In this context,a protocol includes a set of rules defining how the network nodes within the communicationnetwork 109 interact with each other based on information sent over the communication links.Attorney Docket No.: P10355PC00 PatentThe protocols are effective at different layers of operation within each node, from generating andreceiving physical signals of various types, to selecting a link for transferring those signals, to theformat of information indicated by those signals, to identifying which software applicationexecuting on a computer system sends or receives the information. The conceptually differentlayers of protocols for exchanging information over a network are described in the Open SystemsInterconnection (OSI) Reference Model.

[0044] Communications between the network nodes are typically effected by exchangingdiscrete packets of data. Each packet typically comprises (1) header information associated witha particular protocol, and (2) payload information that follows the header information and containsinformation that may be processed independently of that particular protocol. In some protocols,the packet includes (3) trailer information following the payload and indicating the end of thepayload information. The header includes information such as the source of the packet, itsdestination, the length of the payload, and other properties used by the protocol. Often, the datain the payload for the particular protocol includes a header and payload for a different protocolassociated with a different, higher layer of the OSI Reference Model. The header for a particularprotocol typically indicates a type for the next protocol contained in its payload. The higher layerprotocol is said to be encapsulated in the lower layer protocol. The headers included in a packettraversing multiple heterogeneous networks, such as the Internet, typically include a physical(layer 1) header, a data-link (layer 2) header, an internetwork (layer 3) header and a transport (layer4) header, and various application (layer 5, layer 6 and layer 7) headers as defined by the OSIReference Model.

[0045] FIG. 2 is a diagram of the components of the AI-based decision support platform ofFIG. 1, according to one embodiment. By way of example, the platform 101 includes one or morecomponents for analyzing data to determine selection and / or performance of a security / asset orportfolio. It is contemplated that the functions of these components may be combined in one ormore components or performed by other components of equivalent functionality. In thisembodiment, the AI-based decision support platform 101 includes the following modules: an assetallocation module 201, a portfolio construction module 203, a risk management module 205, afinancial planning module 207, a portfolio management module 209, a security selection moduleAttorney Docket No.: P10355PC00 Patent211, an artificial intelligence engine 213, a manager assessment module 215, a unifying indexmodule 217, and a risk of target (RoT) module 219.

[0100] The asset allocation module 201 generates a multitude of dynamic allocation / weightsbased on historic or predicted objective functions (e.g., minimal return of x, max MDD of y, etc.)based on market indices and / or other market data. Additionally, the module 201 can incorporatemacro-economic and social data, thereby providing explainability of how allocations are made(e.g., weighting of measures, macro condition, alternative, etc.). The module 201 has thecapability to change weights or the complete selection criteria, and perform “what if” analyses;this is conducted for selection, tracking or comparative purposes. Further, the module 201provides assessment of historic and ongoing tracking error of allocations. Optionally, the module201 can employ clusters (as opposed to traditional asset class definitions). For example, securityclusters can have behavioral, holding or alternative data similarity (for the duration of the assessedperiod). In effect, the module 201 can satisfy the objective function(s) given evolving marketconditions over a particular time horizon.

[0046] Among other functions, the portfolio construction module 203 generates dynamicportfolio compositions based on historic or predicted objective function (e.g., min return of x, maxMDD of y, etc.). The module 203 utilizes the outputs of the asset allocation module 201 and thesecurity selection module 211. As with the asset allocation module 201, the module 203 providesthe capability to demonstrate how the composition was made (e.g., relationships, macro condition,alternative, etc.), as well as configure the selection criteria and modify weighting parameters. Themodule 203 also supports assessment of historic and ongoing tracking error of portfolios; errorcorrection and penalizing functions are employed for reducing tracking error. Moreover, themodule 203 provides an option to create or optimize portfolios based on clusters, objectivefunction, factors, minus indices, or other criteria.

[0047] With respect to the risk management module 205, this module 205 focuses on theactionable points on a historic, predictive and real-time basis. For instance, the following real-time assessment are supported: portfolio and security sensitivity to factors (e.g., value, trend, etc.);security / portfolio / holdings exposure and give indicators for loss or hedge thresholds and actionAttorney Docket No.: P10355PC00 Patenttriggers; opinions within a given text across news, social media etc. as it relates to the fund,portfolio, holdings, factor, etc. Additionally, the module 205 can provide sentiment analysis onunstructured text into structured data using, for example, Natural Language Processing (NLP).Further, the module 205 utilizes a mathematically stable risk proxy for portfolio comparisons.

[0048] The financial planning module 207 provides accurate forecast of cash flows of assetsunder various scenarios. For example, the module 207 can generate combined cash flow profilesusing a common set of assumptions (e.g., macro, performance measure, etc.). Additionally, themodule 207 can conduct stress and scenario analysis on the combined cash flow profiles, factors,etc. The module 207 utilize multi-step and ML illiquid forecasting models for financial planning.

[0049] The portfolio management module 209 enables management functions relating to aparticular asset, holding, and / or portfolio based on machine learning / deep learning (ML / DL)framework with continuous additions of data, measures and visuals per user requests. Thefunctions of module 209 can be a service that is cloud based, standalone or embedded with anexisting system (e.g., third party system 103).

[0050] The platform 101 provides evaluation, via the security selection module 211, of thebest-performing funds / securities within an asset class, filtered pool or cluster for a given objectivefunction. Such objective function can be based on queries / filters that also combine multiple themes(including pre / post filtering). In one embodiment, the security selection module 211 performsdynamic ranking of securities based on historic or predicted objective functions (e.g., min returnof x, max MDD of y, etc). The module 211 can supplement traditional section criteria ofperformance measures with macro-economic, alternative data and holding data. The module 211supports a capability to explain to the user why the security was selected (e.g., measures, macrocondition, alternative, etc.). The user can specify user-defined performance measures, changeweights, or modify the complete selection criteria. As with the other modules, the security selectionmodule 211 can support selection, tracking or comparative purposes, and provide assessment ofhistoric and ongoing tracking error of the selection.Attorney Docket No.: P10355PC00 Patent

[0051] The artificial intelligence engine 213 interacts with one or more of the various modules201-219 to support the functions of the platform 101. By way of example, the AI engine 213 canexecute the neural network of FIG. 8.

[0052] As depicted, a manager assessment module 215 provides the capability to determineperformance indicators relating to how well, for instance, wealth managers manage their portfolios(financial instruments or any asset).

[0053] Unifying Index module 217 and Risk of Target (RoT) module 219 execute theprocesses of FIG. 4. A Unifying Index (Explainability Index) is generated that captures the assetand / or portfolio's multi-dimensionality and nuances (as captured by any and all performance measures)while remaining easily explainable. In one embodiment, the approach is consistent with thepractitioner's grid approach, but transpose any number of the grid measures as manifolds into aunifying index which can be discussed as a composite or individual components. This proxy isdesignated as the Explainability Index. According to various embodiments, the Explainability Index isbounded between 0 and 1, where 0 is baseline and 1 has a higher degree of variability. Since thereis no limit on the number of measures that be incorporated into Explainability Index, as anextension the Index also gives the practitioner the flexibility to be the decision maker on the choiceof measures, their weights or the final summation process.

[0054] When applied uniformly, Explainability Index gives a consistent assessment acrossassets or portfolios. Where, on a relative basis, if the objective function, benchmark or target hasan Explainability Index of ɳ, then from a desirability perspective the proposed security or portfoliothat is closest to ɳ should be the desirable one. The variability of Explainability Index from thebenchmark or the target objective function is denoted the Risk of Target (RoT). The portfolio orsecurity with a RoT of 0% is the benchmark or the target objective function and the more the RoTvariability the higher the possibility of not hitting the Target.

[0055] Accordingly, a service provider who seeks to provide technical solutions todetermining high performing assets or portfolios face significant technical challenges. There arealso technical challenges with the scenario of producing indices to easily evaluate optimaldecisions. By way of example, these technical challenges include but are not limited to providingAttorney Docket No.: P10355PC00 Patenta unique user experience and interactive platform that comprehensively addresses the problemsand issues described above.

[0056] In one embodiment, to address these technical challenges, the system 100 of FIG. 1introduces an AI-based decision support with the capability to enable one or more practitioners toassess opportunities that satisfy investors’ objectives and circumstances. The unifying measureprovides numerous advantages. First, as noted the Explainability Index, according to oneembodiment, is always a bounded score between 0 and 1; hence a comparison across differentassets or portfolios can be performed on a similar basis. Second, the calculation framework isflexible enough so that as many raw performance evaluation measures can be utilized as desired.Third, based on the categorized input measures, the practitioners can specify their emphasis by auser-defined weighting mechanism which is representative of their evaluation grid preferences.Fourth, the weighted measures can be combined with measures on time variations eitherarithmetically, geometrically or via distribution shifts. There is also a tunable transformationparameter for each real-valued metric which adjustments can be made based on historic orforecasted data across all metrics of similar category. Finally, the module 219 provides aframework for evaluating relative performance via RoT that is a valuable assessment incomparative asset or portfolio evaluation.

[0057] The above presented modules and components of the AI-based decision supportplatform 101 can be implemented in hardware, firmware, software, or a combination thereof.Though depicted as a separate entity in FIG. 1, it is contemplated that the AI-based decisionsupport platform 101 may be implemented for direct operation by respective UE 105. As such,the AI-based decision support platform 101 may generate direct signal inputs by way of theoperating system of the UE 105 for interacting with the applications 107. In another embodiment,one or more of the modules of FIG. 2 and processes of FIG. 4 may be implemented for operationby respective UEs, the AI-based decision support platform 101, or combination thereof. Stillfurther, the AI-based decision support platform 101 may be integrated for direct operation withservices 115, such as in the form of a widget or applet, in accordance with an information and / orsubscriber sharing arrangement. The various executions presented herein contemplate any and allarrangements and models.Attorney Docket No.: P10355PC00 Patent

[0058] FIG. 3 is a diagram of the AI-based decision support platform of FIG. 1 interactingwith various data sources, according to one embodiment. The platform 101 architecturally iscreated in a modular fashion with extensive artificial intelligence-powered data manipulation andanalytical capabilities. In one embodiment, the platform 101 has a flexible architecture and iscloud-based. With respect to securities, such as mutual funds and exchange traded funds (ETFs),the platform 101 can collect robust and cleaned dataset. For instance, the platform 101 cleans andprocesses data from various sources 301 (e.g., market, alternative, client, social, etc.).Additionally, the platform 101 provides the capability to incorporate / append other third-party / unique datasets and advisor holdings. In terms of processing, the platform 101 is modularbased on user (e.g., financial advisor) needs and application programming interfaces (APIs) 303for user ecosystem connectivity. The platform 101 has the capability to select datasets, filters andfeatures for assessment.

[0059] Advantageously, the platform 101 provides a seamless, building block framework withcontinually added depth and breadth. The platform 101 performs an optimized internal processingthat mixes-matches AI techniques / models within the modules best suited for a given objectivefunction. The platform 101 measures tracking error, assess “explainability” of results and abilityto custom train. Module / version release / accretive analysis is readily provided as the platform 101operates from Security to Portfolio to Risk Management (and vice versa) for.Security / Portfolio / Risk Management / Signal / +: measures / metrics (e.g., ratios, Greeks, etc.)presented in visually friendly, non-technical way.

[0060] The platform 101 covers critical portfolio decision facets. For instance, the platform101 monitors real time market signals - jumps, trends, sentiment analysis for sectors, industry. Theplatform 101 also performs security / manager selection based on objective functions - historic orpredictive, as well as asset allocation based on market indices or securities (or security proxies).Optimized or behavioral finance-based custom portfolio construction can be performed using, forexample, traditional methods, clusters or factors. Portfolios can be replicated using other securities(e.g., replicating mutual fund portfolios using ETFs and / or other optimizations). The platform101 can assess custom or uploaded portfolios or transitions with varied holdings. Riskmanagement can be overlaid on market / sector / portfolio / security. Further, the platform 101 canAttorney Docket No.: P10355PC00 Patentperform: Stress testing / Scenario analysis; Factor mapping / sensitivity across all aspects; NLPconnect – news / factors; NLP sentiment – news / factors; and / or Risk Watch – alert / hedging / stoploss.

[0061] The platform 101 provides comprehensive forecasts for financial planning includingincorporating illiquid assets. In this regard, the platform 101 provides: performance / results acrossvarious wealth management modules; amplifies existing advisor methods / tools; continualimplementation of new research within modules; methods / metrics and publishing research; anentirely AI-based with no legacy technical issues; pool of advanced AI-developers and financialmarket specialists; an eco-system - engaging renowned AI and finance experts as advisors.

[0062] FIG. 4A is a flowchart of a process for generating a unifying index by the AI-baseddecision support platform of FIG. 1, according to one embodiment. In one embodiment, the AI-based decision support platform 101 performs the process 400 and is implemented in, for instance,a chip set including a processor and a memory as shown in FIG. 10. As shown, per step 401, theplatform 101 retrieve user-generated preference information relating to selection of the pluralityof the assets or portfolios. The platform 101 also transform the plurality of raw metrics to aplurality of scores that are within a predefined range, per step 403. A composite index or aunifying index (Explainability Index) specifies contributions of each of the performancemeasurements. In step 405, the platform 101 retrieves user-generated preference informationrelating to selection of the plurality of the assets or portfolios . The composite index is generatedbased on the plurality of scores and the user-generated preference information, as in step 407. Theplatform 101 then presents the composite index (e.g., to UE 105a) for determining selection of theplurality of the assets or portfolios (step 409).

[0063] FIG. 4B is a flowchart of a process for generating a relative measure indicator by theAI-based decision support platform of FIG. 1, according to one embodiment. Per process 420,steps 421 and 423, the platform 101 generates another composite index (Risk of Target (RoT))for a target asset or portfolio and determines a relative measure indicator using the compositeindices. The relative measure indicator represents variability from the target asset or portfolio.The platform 101, per step 425, presents the relative measure indicator, for example, to a graphicalAttorney Docket No.: P10355PC00 Patentuser interface (GUI) of UE 105a. The RoT can be used to rank assets or portfolios, assesscomposite or component deviations, evaluate point-in-time, relative or trends, and will be helpfulin optimization decisions as well. Moreover, RoT has application for more securities in thefinancial sector (both liquid and illiquid), incorporate alternative and holding data (e.g., trackrecord, expense ratios, manager turnover, turnover ratio, concentration ratio, etc.) and itseffectiveness for asset allocation, portfolio construction, risk management and asset planning.Further details of process 400 are as follows.

[0064] Performance measures are constructed by forms of manipulation of the same underlierthat is time series of price. The most primitive performance measure would be measuring totalreturn for a period under consideration, say T.

[0065] where is the value or price of the ith underlier at time t and its (total) returnfrom t to T. This novice performance measure is agnostic to path traveled0 to T. This noviceperformance measure by dividing it by the range for period T, gradually introducing volatility ofthe path traveled. This is called Gain to Range ratio (GRR):

[0066] Anotherof the path traveled,i.e., Sharpe ratio.

[0067] Process 400 can utilize any conventional portfolio performance measures; some ofthese performance measures are relative, and thus need a designated benchmark. Further,according to certain embodiments, the performance measure calculation can be based on a wholeperiod of historical / forecasted data or based on a rolling window analysis within thehistoric / forecasted data. To simplify assessments, the various performance measures can begrouped by categorizing them as return-targeting, volatility-targeting, drawdown-targeting,Attorney Docket No.: P10355PC00 Patentvariation-targeting, etc. As the category name suggests, some metrics are focusing on volatility orsome form of it; some are focusing on drawdown or variation of it; and some are trying to look atthe return from its empirical distribution perspective by focusing on other moments or somenonlinear characteristics. The categorization and compositions presented herein are for illustrativepurposes.

[0068] With respect to the category of return-targeting, for example, the performance measuresare related to return – e.g., including total return, average return, or excess return with respect tointerest rate or inflation rate, and target return. In some cases returns be adjusted for the time valueof money. Nevertheless no other factors such as volatility and drawdown come into calculation ofthe performance measure.

[0069] Regarding volatility-targeting, this category of measures relate to the volatility ofreturns – e.g., including Sharpe ratio, Sortino ratio, and proxies such as Bollinger band crossingsand absolute returns proxy for volatility. Sharpe ratio is the widely-used performance metric inportfolio management due to the fact that it computes a single number which takes into accountboth the first and the second moments of the returns distribution. Improvements on the Sharperatio include Information ratio and Sortino ratio. Information ratio is a slight modification ofSharpe ratio by replacing the risk-free returns with a reference / benchmark portfolio. TheInformation ratio measures a portfolio’s ability to generate better performance than a benchmarkportfolio. The Sortino ratio considers only harmful volatilities, and replaces the standard deviationwith a downside deviation that only takes into account returns below a certain threshold.

[0070] An important aspect of portfolio management is the measurement of drawdown risk.Drawdowns are different from returns in that they capture the psychological effect of long-termfinancial distress by accumulating consecutive losses, therefore providing insights that are notcaptured by looking only at the returns. Some common measures of drawdown risk includemaixmum drawdown, average drawdown, and conditional drawdown-at-risk.

[0071] In the variation-targeting category, performance measures relate to statistical propertiesof the return distribution. If the underlier is benchmark related, correlation and R-squared comeinto the play as it is essential to know how correlated the underlier is with respect to its benchmark.Attorney Docket No.: P10355PC00 PatentHigher moments statistical properties such as skewness and kurtosis would be considered as well.The Jarque-Bera test, a goodness-offit test of whether sample data have the skewness and kurtosismatching a normal distribution, indicating how far the underlier performance is from a normaldistribution.

[0072] The above described measures can be computed based on a whole period of historicalor forecasted data, according to various embodiments. These approaches can be extended to arolling window analysis within the historic or forecasted data, which allows us to extractinformation about the time variation and consistency of the performance measures for thatportfolio. Assuming use of historical or forecasted returns data of for a portfolio of size T, awindow size H and a forward looking step size of h can be utilized, yielding: returndata segments of length H within the (univariate) full sample data ofnumberrounded down, towards zero, to the nearest integer. Each piece of data is denoted.case of a fixed period, the time series from the last window are used to calculateall performance measures for that window, without examining variation across time. In the caseof variation across time, performance measures for each segment can be calculated and time seriesfor each of them are generated. Having time series of each performance measure, a single metriccan be computed to assess the variation of each measure across time. By way of example, themetric to assess variation of any measure across time can be Sharpe ratio (or any of the otherdescribed modifications of Sharpe ratio). No matter which measure is under consideration, theSharpe ratio of times series of that measure can be calculated -- which could be time series ofSharpe ratio itself.

[0074] Process 400 utilizes a unifying measure that incorporates any combination of theselected performance measures, while retaining their properties. For example, consider a list of Kraw measures for portfolio performance, which can take values in the entire real space. This isdenoted by . The first step in constructing a unifying measure is to transform allAttorney Docket No.: P10355PC00 Patentmeasures into a unique and finite range. A transformation is applied that maps every metric to anumber between 0 and 1.

[0075] With respect to the transformation technique (which can be one-step or two-steptransformation), the following notations are defined:: raw value of the performance measure e.g. Sharpe ratio, Calmar ratio, etc.: line ar transformed value of (two-step transformation): nonlinear transformed value of (one-step transformation) or (two-steptransformation) :adjusted value of if needed

[0076] In the one-step transformation, in order to transform any performance metric, m, to bebounded between 0 and 1, the following sigmoid transformation can be used:

[0077] The problem is thatnot have the same range and density, somemove in a tight range and others in a wider range, and the nonlinear transformation sigmoid wouldnot be able to reflect that. One approach is to apply the scaled version of this transformation, thatis:where γ is a tunable hyperparameter.

[0078] By way of example, the result of this transformation on real values between -5 and 5for different values of γ = 1, 2, 3, 4, 5 reveals that it is symmetric around (0, 0.5); and as the rawmetric becomes larger in magnitude it quickly converges to the extreme values of 0 and 1. If thissimple transformation is applied to some of the performance measures it may just cover the rangepartially. For example, for the maximum drawdown, volatility and batting average, thetransformed value would be between 0.5 and 1.0. Tuning the hyperparameter γ, would not be easyas the index should be easily explainable. Moreover, a bias factor needs to be introduced to theAttorney Docket No.: P10355PC00 Patentscaled sigmoid transformation to make sure (0, 0.5) coincides with the mean or median of theperformance measure under consideration.

[0079] The one-step transformation advantageously is simplistic, and can be applied in certainembodiments. To overcome the issues in the one-step transformation, process 400 can utilize atwo-step transformation. In the two-step transformation, first a linear mapping transforming α × 100% and (1 − α) × 100% of a performance measure to the corresponding α × 100% and (1 − α) × 100% of the sigmoid function is utilized. It is noted that α × 100% and (1 − α) × 100% of a performance metric can be obtained numerically. Alternatively, α × 100% and (1 − α) × 100% of the sigmoid function can be computed analytically as follows:and for can do the same calculation to obtain:Attorney Docket No.: P10355PC00 PatentDenoteto see thatDenote α × 100% and (1 − α) × 100% of a performance metric, zα and z1−α respectively that is

[0080] Linearand x1−α or equivalently −a and a.The slope for this linear transformation is given by

[0081] Having the slope, theis obtained:

[0082] Having the linear transformation, is passed to the sigmoid function to yield:Attorney Docket No.: P10355PC00 Patent

[0083] It is noted that in these operations, 0 < α < 0.5. In the two-step transformation, thepoint (0, 0.5) coincides with the mean of the performance measure distribution.

[0084] The platform 101, in one embodiment, can adaptively learn the transformationparameters based on information distances between the data set under such transformations, andprovide more balanced raw metrics for the final computation of the unifying measures.

[0085] Process 400 at this stage can compute the Explainability Index. That is, after thesetransformations, a list of measures are bounded between 0 and 1. The lower valuesindicate higher preference may indicate undesirable instances. For all performance measureswhere larger values are desirable, e.g. Sharpe ratio, Calmar ratio, etc., the following adjustmentcan be made:

[0086] On the other hand, for those where large values are not desirable such as volatility,drawdown, etc., no adjustment is needed:

[0087] Therefore each portfolio canby the high dimensional vectorwhich lives inside a unit hyper-cube. According to various embodiments, theis calculated based on all the above information and some user generated preferenceparameters. The intuition behind the preference parameters is that some practitioners tend to valuethe return-chasing metrics more while others value the drawdown-based or volatility focusingmetrics more. To tune the unifying index towards different investor’s tastes, a set of non-negativeweight parameters can be defined for each of the main categories of metrics. Defaultvalues for these weights would be non-informative choice of . But in general, these weights areeither directly given by the practitioner based on his / her preferences, objective, or risk appetite orcomputed from the ranking of such preferences. Combining these weights with the transformedscores yields the following score (which serves as a weighted distance measure between the originand the current portfolio in the portfolio metric hyper-cube):Attorney Docket No.: P10355PC00 Patent

[0088] It is noted that the higher weights will put more emphasis on that specific measure inthe overall score, and the Explainability Index is always bounded between 0 and 1, with 0indicating the highest preference. To enhance the comprehensivity of the broad range of proposedmeasures, such measures are categorized into the four described distinct groups: return targeting,volatility targeting, drawdown targeting and variation targeting. The platform 101 provides arobust framework that is designed to employ none or any number of such categories. Accordingto one embodiment, the measures within each category are equally weighted to produce thecategory score and the four higher-level category scores are weighted using user risk preferencesto produce one final score or the Explainability Index.

[0089] Since it is possible that the arithmetic mean can be biased towards certain categoriesof measures, for which the scores tend to be closer to one; flexibility for a geometric mean typecalculation procedure is introduced, utilizing the weights associated with user preferences for allfour categories, represented as the geometric Explainability Index. The exact formula is presentedbelow, similar to Explainability Index above:

[0090] According to one embodiment,the ExplainabilityIndex is multiplied by the distance shift factor. The distribution shift factor intends to measure thedegree of distribution shift within the data. Following previous notations, each data segmentwill correspond to a return distribution that is different from the full distributionincludes all the data points. Information distances indicate how close two distributionsare from one another. Some well-known information distances and divergences are: (a) Kullback-Leibler divergence, (b) Jensen-Shannon divergence, and (c) squared Hellinger distance, which isAttorney Docket No.: P10355PC00 Patenta special case of a more general family of f-divergences. For two probability distributions P andQ with densities denoted by f(x) and g(x) respectively, the squared Hellinger distance is definedas:

[0091] It is notedsymmetric and boundedbetween 0 and 1. As a known special case, for two normal distributions and, the squared Hellinger distance takes the following

[0092] Based on thisdistance can becalculated from the full distribution for each data segment:

[0093] Empirically, thesample data can be used to computethis distance or use histogram approximations to compute a discrete squared Hellinger distance.Alternatively an empirical squared Hellinger distance estimator can be used for univariatedistributions; this uses only the two data samples and is guaranteed to converge as data sizeincreases. Once all the distance metrics for each data segment are computed, an average distancemetric is computed -- which reflects how much the return distribution has shifted within the wholehistoric period:

[0094] The distribution shift is also bounded between 0 and 1.The closer this distance is to zero, the more consistent the portfolio return distributions are, whichAttorney Docket No.: P10355PC00 Patentis generally more desirable. The next step is to combine the distribution shift distance with theExplainability Index (EI). For portfolios that are highly volatile in-sample the distribution shiftdistance is supposed to amplify the EI -- i.e. make it go toward one. That implies that themultiplication of EI by dH cannot be done, as it would significantly dampen the value or improveperformance. For example, EI of 0.2 and dH of 0.1 would yield 0.2 × 0.1 = 0.02. The newly obtainedvalue is supposed to be equal to or higher than EI. Process 400 reflects both values through 0.5(same as subtracting it from one) by multiplying them and reflecting the product back -- that is:1 − (1 − 0.2) × (1 − 0.1) = 0.28This obtained value is higher than EI as expected.

[0095] Therefore distributional Explainability Index is obtained as follows:dExplainability Index = 1 − (1 − Explainability Index) × (1 − dH)

[0096] The final output score is defined as a distributional Explainability Index and boundedbetween 0 and 1, with 0 indicating the highest preference.

[0097] When applied uniformly, the Explainability Index gives a consistent assessment ofassets or portfolios that can also be used on a comparative basis. For example, the ExplainabilityIndex of the objective function, benchmark or target can be compared with the Explainability Indexof the asset or portfolios. As previously noted, this variability of Explainability Index of the assetor portfolio from the benchmark or the target objective function is defined as the Risk of Target(RoT); wherein, the portfolio or security with a RoT of 0% is the benchmark or the target objectivefunction and the more the RoT variability the higher the possibility of not hitting the Target.

[0098] Theaor on the objectivefunction. For example, for index tracking Exchange-Traded Funds (ETFs) any variation in Risk ofTarget (RoT) maybe not be desirable, but a low Risk of Target (RoT) on portfolios may implybetter performance than the target.Attorney Docket No.: P10355PC00 Patent

[0099] According to one embodiment, in the calculation of an Explainability Index, process400 need not incorporate correlation. Performance measures within each category are highlycorrelated; however, due to averaging, the correlation gets subdued or eliminated. Crosscategories, in general, are not supposed to be correlated but if there is any correlation, it would notbe substantial and would not have a big impact on the index considering correlation crosscategories also smoothed out. Another rationale for not considering the correlation factor has todo with yet introducing it as another hyperparameter that could have undesirable instability andcreate sensitivity in the calculation of the Explainability Index which requires tuning each time.

[0100] In the calculation of an Explainability Index, process 400, according to variousembodiments, need not normalize the performance measures or categories. The lineartransformation already includes elements of centering and scaling, resulting in a negligible effecton whether the data is normalized. The linear transformation shows to be preferable to thenormalization given that it allows to account for percentage thresholds on the sigmoidtransformation. The variation targeting category has measures that show greatest skewness overall,leading to most points concentrating around 0 or 1.

[0101] It is contemplated that platform 101 can provide several extensions to the describedframework, including various categorization or weighting mechanisms on input measures to reflectinvestor risk preferences, methods to incorporate variation of measures over time, geometricsummed version (defined as “gExplainability Index”), as well as incorporating distributional shiftswithin a “dExplainability Index.” Additionally, alternative tuning of the score transformationmethod in the calculation procedure can be utilized to balance different classes of measures and togenerate scaled Index values.

[0102] FIGs. 4C and 4D are flowcharts of processes for utilizing machine learning models forasset planning by the AI-based decision support platform of FIG. 1, according to variousembodiments. Process 430 shown in FIG. 4C provides a two-step prediction model. Cash flowdata associated with one or more assets is received (as in step 431). Process 430 then, per step433, applies a best-fit model to the cash flow data. A regression model is applied, per step 435, toAttorney Docket No.: P10355PC00 Patentthe cash flow data to predict error term. Next, process 430 outputs prediction data based on thebest-fit model with the error term prediction (per step 437).

[0103] Additionally or alternatively (shown in FIG. 4D), process 440 provides various models(e.g., a Cash Flow Prediction Model (CFPM)) and involves receiving historical performance dataand economic data associated with one or more illiquid assets (as in step 441). In step 443, amachine learning model is trained using the historical performance data to predict one or morefeatures including cumulative contributions, distributions, and net asset value (NAV) percentagechanges. In step 445, process 440 incorporates regime-based transition probabilities into themachine learning model to account for different economic stages. Next, process 440 generatescash flow forecasts and stress testing output for the illiquid assets based on the trained model (asin step 447).

[0104] Further details of the above processes are described as follows. By way of illustration,the data processed by platform 101 relates to financial data, including data of an asset class (i.e., asingle asset or commingled fund) as well as macroeconomic data.

[0105] Data providers are not always able to obtain all available cash flow data for a givenasset or fund. As a result, such providers include whatever information they can collect – the datais not necessarily the same information type across funds. Depending on the asset or fund, it isrecognized that providers could collect any of the following parameters: commitment,contributions, distributions, NAV (Net Asset Value), RVPI (Residual Value to Paid-In-Capital),% of capital called in, TVPI (Total Value to Paid-In-Capital), etc. To obtain standardized andmore usable information, data is converted into ratios, according to one embodiment.

[0106] Missing values are encountered at many points during the life of the fund, e.g., thebeginning, middle, or end of the life of the fund. Sometimes missing values could also beencountered periodically; for example, only semi-annual information are obtained for some fundsinstead of quarterly. To utilize the greatest possible number of funds as well as the lowest possiblefrequency (quarterly), the missing values are filled for the fund. There is a distinction between amissing value “today” and a missing value in the past. For past missing values, interpolation canbe performed between two data points, but for present missing values interpolation cannot beAttorney Docket No.: P10355PC00 Patentutilized, as “future” datapoints are not available. Accordingly, missing values fill in a rollingwindow basis, where past values could be filled in using linear interpolation, cubic spline or fillforward while present values will always be filled in using fill forward.

[0107] For most funds, the data resembles a step function. Funds do not have contributionsand distributions every single quarter. This makes the cumulative contributions and distributionslook like a step function. Step functions are mathematically harder to work with regardless of themodel utilized. In the case of the deterministic model, not having distributions for a quarter makesthe bow calculation unstable. Thus, according to various embodiments, a Heat Equation isemployed to smooth the cumulative contributions and distributions. When aggregating multiplefunds to train a model, it is observed that the behavior of the fund heavily relies on the age of thefund (i.e., “vintage stacking”). The expected behavior of, for example, a 3 year old fund is vastlydifferent than the one of a 10 year old fund. Model formulations will train a different model foreach fund age. These can be built into the models as options for forecasting and stresses.

[0108] Similar to vintage stacking, fund’s behavior can be determined by its performance withrespect to its peers, so that formulation can also train a different model with respect to relativeperformance. The models can be constructed as options for forecasting and stresses.

[0109] Macroeconomic data is necessary to train the models as well as for risk management --- e.g., stress to assess the sensitivity of the cash flows to macro indicators.

[0110] It is noted that traditionally, there are hundreds of indicators representing the state ofthe economy. Since this can be unmanageable, the platform 101 utilizes a classification scheme togroup the indicators based on related economic aspects. A substantial number of indicators shouldbe included within each group to properly represent an economic sector. It is important to makethe indicators stationary.

[0111] It can be complicated to make sense of the movements of hundreds of indicatorsunderlying the macroeconomic data. Most of the data is also highly correlated. Thus, the groupsare formulated to decrease the number of indicators into a manageable number as well as to reduceredundancy and noise within the data. This process leaves us with a macroeconomic factor foreach group.Attorney Docket No.: P10355PC00 Patent

[0112] It is recognized that historical sections of the economy can also be classified intodifferent regimes. Regimes allow the model to learn different behaviors of the cash flows duringdifferent stages of the economy.

[0113] The assessment of the asset or fund hinges on the relative range of an individualperformance measure’s value of the asset or fund. This relative range when looked at as a gridwith the selected measures, and associated weights, facilitates the practitioners’ preferences inselecting the asset or fund. This is usually supplemented by some form of backtesting and / orsimulation giving the practitioner more comfort around the choice of the measures, the weights,and consequently the asset or fund being evaluated.

[0114] Financial performance measures are largely derived from the manipulation of theunderlying cash flows. Traditionally, many measures have been proposed and practitioners defineand formulaically use them with degrees of freedom. With so many measures to choose from thefinal assessment or communication, making comparisions among assets or funds difficult(especially since the scales, methods, etc. can be very different).

[0115] Accordingly, as earlier described an Explainability Index (EI) is introduced, providinga simple and effective way to incorporate all desired individual measures into a composite index.This makes assessing and communicating relative assessments and drivers of performanceuniform, simple, and scalable. EI unifies all measures and gives a composite index between 0 - 1(with 0 being better). Thereafter, Risk of Target (RoT) highlights the divergence of individualmeasures from the Target (e.g., benchmark). The methodology can incorporate any number ofmeasures, including alternative data -- e.g., track record, leverage, ownership, etc. for assessmentsat inception or post-commitment.

[0116] In terms of forecasting, for deterministic models, certain assumptions are made totransform the following parameters: growth rate (g), contribution rate (cr), yield (y), and bow (b).The growth rate can be derived directly as a geometric mean, for example, a 5% annual growthrate equals a 2.47% growth rate per half year or 1.23% per quarter. Contribution rate can be derivedfrom deterministic equations plus adding certain new conditions, such as the sum of contributionsAttorney Docket No.: P10355PC00 Patentof each quarter has to be equal to the contribution of the entire year. A closed form solution canbe solved for the contribution rate; however, there is no close formula for bow or yield.

[0117] Manipulating the contributions formula to make it into semi-annual steps:

[0118] The following

[0119]

[0120] For growth rate, thecanAttorney Docket No.: P10355PC00 Patent

[0121] There is no close formula for distribution rate, but the same assumption can be madeas the contribution rate:

[0122] According to variousdata of each fund / company can beutilized to obtain the best historical fit assumptions. The deterministic model formulas can bemanipulated to obtain the fitted parameters, including quarterly and simi annual data.

[0123] Utilizing historicalthe following:

[0124] The best fit forN AV formula:

[0125] From the distribution rate formula, two variables b and L are present. A value for Lcan be selected, whereby b is solved for the selected value.A market line is generatedsame process as with best-fit canleveraged, but instead of applying the formulas to the fund’s historical cash flows, they are appliedto the market lines. Because multiple vintages exists, multiple set of parameters are provided forwhich an average can be determined to obtain the final expected parameters for each fund basedAttorney Docket No.: P10355PC00 Patenton their age. This allows for benchmarking as well as for making better assumptions especially forfunds in their early life.k stands for available market lines

[0126] Similar to the market line based models, Regime based models can be constructed bycomputing the market line parameters during each regime based on the probability of being in agiven regime. These parameters can then be combined based on the probabilities of the currentregime. Different regime parameters can be used for the multi-step.

[0127] The platform 101in those historical parameters asthese yield better results. The past historical quarters are employed, whereby the best-fit methodis used to obtain a set of parameters, the “future” quarters are then used as well as the best-fitmethodology -- obtaining crx,t and cry,t. Next, the difference is taken: dcrt = crx,t – cry,t.Subsequently, using the historical cash flows and any performance metric, a regression algorithmis utilized to predict the change in the parameters. Finally, the predicted change in the parameterdcrt is added to crx,t to obtain . The process is repeated for the remaining two parameters; suchparameters can applied to the deterministic model to obtain the final predictions.Attorney Docket No.: P10355PC00 Patent

[0128] Learning-based models assume that cash flow profiles can be generated based onlearned behaviors. In general, the deterministic model captures the profile, but has more of a fixedpath, where the actual performance may not be as smooth. A two-step prediction model thataddresses this problem is utilized. The best-fit model is applied to the data and subtracted by theactual cash flows. The remainder would result in an error term of the deterministic modelprediction which is not deterministic. A regression model is then applied to predict the error term.The prediction of the best-fit together is combined with the non-deterministic prediction to obtainthe final prediction result.

[0129] Turning now to the Cash Flow Prediction Model (CFPM), this model departs from thedeterministic model.

[0130] The CFPM model aims to predict the cumulative contribution, cumulative distribution,and NAV values during the prediction horizon at certain quarters. By way of example, predictionof the last quarter and the middle quarter is selected for the prediction horizon. For the remainingvalues, a cubic interpolation followed by a heat equation smoothing is utilized.

[0131] At each prediction, error terms are introduced from the model itself, given the reasonwhy not to predict every quarter but only for chosen quarters. Also, on a quarterly basis, the datatends to be more noisy. There are many quarters with no distributions or contributions while somequarters present big jumps in the data. Predicting every [X] quarters allows for those jumps to bediluted and better capture the expected performance of a fund in the short to medium term.

[0132] As for the regression itself, performance indicators are included with macroeconomicdata and used a Random Forest Regressor as default; however, it is noted that any regression modelcan be used. The following features are considered: cumulative contributions, cumulativedistributions, and NAV -- given that future performance is directly correlated with it (e.g., ifcumulative contributions are close to 100%, future contributions are expected to be of lowmagnitude, and if NAV is close to 0%, large distributions are not expected). Additionally, theslope of the previously mentioned features is determined; such information reveals not only on themagnitude of previous performance but also on the rate at which they are changing. The formulafor the slope is as follows:Attorney Docket No.: P10355PC00 Patent

[0133] As some models outperform others during certain market conditions or for certainfund types, the model predictions are combined to reduce the variance error. As a start, the models’predictions are equally weighted. Sequentially, the model predictions are weighted based on theprevious realized prediction error based on the funds age l. Reduction of the dollar value in theprediction (which error could take positive or negative values) is elected. The realized predictionerror is calculated as follows:where i stands for model i,for age l, Cpi,l,f,t for model icontribution prediction for age l, fund l and Cal,f,t for the funds actual prediction.

[0134] Then the positive and the negative model errors are weighted based on their inverseerrors.

[0135] Weights for positive and negative individual models are as follows:Attorney Docket No.: P10355PC00 Patent

[0136] The final weight isweight and the previous weight.

[0137] The above described model extends the traditional primary reliance on thedeterministic model and offers a variety of financial models for relating different models to betterfit different fund profiles or asset structures. The final selection of a model depends on thenuances / features of the asset or fund.

[0138] With respect to risk management, traditional approaches for illiquid assets centeraround implementing a few simplistic what-if scenarios to the deterministic model inputparameters (e.g., increase commitments, decrease distributions, etc.). Such approaches have a fewissues: a) it relies on the deterministic model (which has a fixed profile and may not fit allasset / fund structures; b) is ad-hoc and largely driven by anecdotal scenarios for the inputparameters; and c) if the ad-hoc illiquid results are combined with structured stresses on liquidassets then the overall asset planning approach is very arbitrary.

[0139] The platform 101 provides a framework that allows for assessing the sensitivity of cashflows using traditional macroeconomic indicators, regimes, historical scenarios, and so on withinthe expanded financial model offerings. Conducting structured stress testing and simulations withbetter-fitted financial models offers a more robust and uniform asset planning framework (acrossliquid and illiquid assets).

[0140] The macroeconomic indicators are regressed with the fund parameters andperformance to ascertain the relationships individually, in part as categories, as vintages and so on.This permits relating the stresses on the macroeconomic indicators on the parameters or the cashflows via the relationships. The complexity lies in stabilizing the relationship sensitivity tocategories, vintages, performance profiles and fund life. This includes the covariance matrix as itAttorney Docket No.: P10355PC00 Patentevolves over time. The stressed parameters allow the generation of stressed cash flow forecastsand related performance measures. The stresses can also be via the regime models to account forthe difference in the expected change of macroeconomic indicators during different regimes.

[0141] According to various embodiments, the macroeconomic signals are reduced from atraditionally large set. Without such reduction, there would be too many regressors per number ofdata points, thereby causing overfitting the regression and leading to fitting irrelevant indicatorsinto the regression. To obtain a meaningful stress test, the proper steps are to apply reduction ofthe macroeconomic variables to a more manageable number, and then applying the regression. Theoriginal data can then be stressed tested and through the inference in the covariance matrix, theexpected change can be calculated for the reduced measure to then obtain the expected change intothe parameter itself.

[0142] Because a relatively fewer indicators are utilized, the indicators covariance matrix isapplied to determine their expected changes given particular assumptions. To determine theeffectiveness of the covariance matrix, the historical macroeconomic data is passed for somesignals, while others are inferred.

[0143] With respect to a deterministic based model, a regression is fitted to determine therelationship within the current macroeconomic data to the expected average change of thedeterministic model parameters. Then the change is implemented into the historical best-fitparameters. Finally, the deterministic model is utilized to determine the final expected changes inthe fund cash flows given the current state of the economy.

[0144] Under the learning-based approach, the relationship between the macroeconomic dataand the cash flows are directly determined. The macro data to the cumulative contributions,distributions and NAV percentage change during the stress period is regressed. Additionally, thelast quarter's values and the middle one are regressed. Sequentially, the percentage changes to areapplied to stressed performance to then interpolate within the stress values.

[0145] The described models provide for a structured asset planning approach across allassets. The new financial models are specifically tailored for illiquid assets so there are choices toAttorney Docket No.: P10355PC00 Patentfit the asset or fund nuances / features. Such models extend traditional liquid asset stress tests toilliquid assets (for uniform stress testing across the portfolio).

[0146] FIGs. 5A-5H are diagrams of a graphical user interface (GUI) relating to datamanagement processes performed by the AI-based decision support platform of FIG. 1, accordingto one embodiment. GUIs 501, 503, and 505 permit the user to specify the different types ofsources of data to be collected. Such sources can be identified and selected based on asset types,macroeconomic factors, media, or even proprietary data. Via GUIs 507 and 509, the platform 101can perform predictions on the selected data sources using various models: historic, predicted,stress testing, and / or backtesting for example. According to one embodiment, GUI 511 supportsdrilling down to permit user input of various aspects of mutual funds. GUI 513 illustrates theclustering feature of platform 101. Additionally, with GUI 515, different views of the dataprocessing of clusters are supported.

[0147] FIGs.6A and 6B are diagrams of a GUI relating to asset allocation processes performedby the AI-based decision support platform of FIG. 1, according to one embodiment. Namely,through GUIs 601-605, different filters can be applied regarding how asset allocation can beexecuted. GUI 607 presents performance data associated with a selected asset or cluster of assets.

[0148] FIGs. 7A-7D are diagrams of a GUI relating to security selection processes and indexgeneration processes performed by the AI-based decision support platform of FIG. 1, according toone embodiment. GUIs 701 and 703, by way of example, provides specification of various datafilters for an asset. GUI 705, as in FIG. 7D, provides a section showing categories of performancemeasures along with the associated weights.

[0149] FIG. 8 illustrates an example neural network 801 (e.g., an example of the AI engine211 implementing a machine learning model) that has an architecture including an input layer 803comprising one or more input neurons 805, one or more hidden neuronal layers 807 comprisingone or more hidden neurons 809, and an output layer 811 comprising one or more output neurons.In one embodiment, the architecture of the neural network 801 refers to the number of inputneurons 805, the number of neuronal layers 807, the number of hidden neurons 809 in the neuronallayers 807, the number of output neurons, or a combination thereof. In addition, the architectureAttorney Docket No.: P10355PC00 Patentcan refer to the activation function used by the neurons, the loss functions applied to train theneural network 801, parameters indicating whether the layers are fully connected (e.g., all neuronsof one layer are connected to all neurons of another layer) or partially connected, and / or otherequivalent characteristics, parameters, or properties of the neurons 805 / 809 / 813, neuronal layers807, or neural network 801. Although the various embodiments described herein are discussedwith respect to a neural network 801, it is contemplated that the various embodiments describedherein are applicable to any type of machine learning model 109 that can be migrated betweendifferent architectures.

[0150] In one embodiment, the progressive path migrates an old architecture of a machinelearning model 109 into a new architecture by incrementally adding and removing single neuronsor neuronal layers, or smoothly changing activation functions in a fashion which does not affectperformance of the machine learning model 109 by more than a designated performance changethreshold. For example, a user may wish to migrate a machine learning model 109 from anarchitecture that has three hidden neuronal layers 807 with four hidden neurons 809 in each layerto a new architecture that has four hidden neuronal layers 807 with four hidden neurons 809 each.The machine learning model 109 has been trained using the old architecture for a significant periodof time. To advantageously preserve the training already performed and maintain modelperformance at a target level, the system 100 can construct a progressive path with four steps thatincremental adds one hidden neuron 809 to the new neuronal layer 807 at each step until the fullnew neuronal layer 807 is added. In other words, while the machine learning model 109 of themachine learning system 107 is being trained, a new technical solution or architecture may bediscovered that can provide improvements to the machine learning model 109 or system 107. Theninstead of replacing the old system architecture in a cut-off fashion, the system 100 can constructincremental steps that can be used to progressively migrate the existing trained machine learningmodel 109 to avoid catastrophic degradation of the trained machine learning model 109’sperformance.

[0151] In one embodiment, while the progressive migration is being done, the training processcontinues. In this way, the newly added neurons learn relatively quickly their new roles in themachine learning model 109 as their context environment consists of neuronal layers 807 whichAttorney Docket No.: P10355PC00 Patentalready know their jobs (e.g., neuronal layers 807 with neurons 809 that have undergone at leastsome training). After migration the resulting machine learning model 109 has incorporated expertknowledge from the old architecture, but has a new architecture, new technologies incorporated,and / or the like which can potentially improve the performance and learning of the machinelearning model 109 in the future. Accordingly, the embodiments of the system 100 describedherein provide technical advantages including, but not limited to, providing long-lived machinelearning systems 107 that can be trained better while incorporating new advances in machinelearning technologies (e.g., neural network technologies).

[0152] The processes described herein for providing decision support may be advantageouslyimplemented via software, hardware, firmware or a combination of software and / or firmwareand / or hardware. For example, the processes described herein, may be advantageouslyimplemented via processor(s), Digital Signal Processing (DSP) chip, an Application SpecificIntegrated Circuit (ASIC), Field Programmable Gate Arrays (FPGAs), etc. Such exemplaryhardware for performing the described functions is detailed below.

[0153] FIG. 9 illustrates a computer system 900 upon which various embodiments of theinvention may be implemented. Although computer system 900 is depicted with respect to aparticular device or equipment, it is contemplated that other devices or equipment (e.g., networkelements, servers, etc.) within FIG. 9 can deploy the illustrated hardware and components ofsystem 900. Computer system 900 is programmed (e.g., via computer program code orinstructions) to provide decision support as described herein and includes a communicationmechanism such as a bus 910 for passing information between other internal and externalcomponents of the computer system 900. Information (also called data) is represented as a physicalexpression of a measurable phenomenon, typically electric voltages, but including, in otherembodiments, such phenomena as magnetic, electromagnetic, pressure, chemical, biological,molecular, atomic, sub-atomic and quantum interactions. For example, north and south magneticfields, or a zero and non-zero electric voltage, represent two states (0, 1) of a binary digit (bit).Other phenomena can represent digits of a higher base. A superposition of multiple simultaneousquantum states before measurement represents a quantum bit (qubit). A sequence of one or moredigits constitutes digital data that is used to represent a number or code for a character. In someAttorney Docket No.: P10355PC00 Patentembodiments, information called analog data is represented by a near continuum of measurablevalues within a particular range. Computer system 900, or a portion thereof, constitutes a meansfor performing one or more steps of the processes described herein, including that of FIG. 4.

[0154] A bus 910 includes one or more parallel conductors of information so that informationis transferred quickly among devices coupled to the bus 910. One or more processors 902 forprocessing information are coupled with the bus 910.

[0155] A processor (or multiple processors) 902 performs a set of operations on informationas specified by computer program code related to providing decision support. The computerprogram code is a set of instructions or statements providing instructions for the operation of theprocessor and / or the computer system to perform specified functions. The code, for example, maybe written in a computer programming language that is compiled into a native instruction set ofthe processor. The code may also be written directly using the native instruction set (e.g., machinelanguage). The set of operations include bringing information in from the bus 910 and placinginformation on the bus 910. The set of operations also typically include comparing two or moreunits of information, shifting positions of units of information, and combining two or more unitsof information, such as by addition or multiplication or logical operations like OR, exclusive OR(XOR), and AND. Each operation of the set of operations that can be performed by the processoris represented to the processor by information called instructions, such as an operation code of oneor more digits. A sequence of operations to be executed by the processor 902, such as a sequenceof operation codes, constitute processor instructions, also called computer system instructions or,simply, computer instructions. Processors may be implemented as mechanical, electrical,magnetic, optical, chemical, or quantum components, among others, alone or in combination.

[0156] Computer system 900 also includes a memory 904 coupled to bus 910. The memory904, such as a random access memory (RAM) or any other dynamic storage device, storesinformation including processor instructions for providing real-time data analysis to supportdecision making. Dynamic memory allows information stored therein to be changed by thecomputer system 900. RAM allows a unit of information stored at a location called a memoryaddress to be stored and retrieved independently of information at neighboring addresses. TheAttorney Docket No.: P10355PC00 Patentmemory 904 is also used by the processor 902 to store temporary values during execution ofprocessor instructions. The computer system 900 also includes a read only memory (ROM) 906or any other static storage device coupled to the bus 910 for storing static information, includinginstructions, that is not changed by the computer system 900. Some memory is composed ofvolatile storage that loses the information stored thereon when power is lost. Also coupled to bus910 is a non-volatile (persistent) storage device 908, such as a magnetic disk, optical disk or flashcard, for storing information, including instructions, that persists even when the computer system900 is turned off or otherwise loses power.

[0157] Information, including instructions for providing real-time data analysis to supportdecision making, at least in part, on analysis of collected information, is provided to the bus 910for use by the processor from an external input device 912, such as a keyboard containingalphanumeric keys operated by a human user, a microphone, an Infrared (IR) remote control, ajoystick, a game pad, a stylus pen, a touch screen, or a sensor. A sensor detects conditions in itsvicinity and transforms those detections into physical expression compatible with the measurablephenomenon used to represent information in computer system 900. Other external devicescoupled to bus 910, used primarily for interacting with humans, include a display device 914, suchas a vacuum fluorescent display (VFD), a liquid crystal display (LCD), a light-emitting diode(LED), an organic light-emitting diode (OLED), a quantum dot display, a virtual reality (VR)headset, a plasma screen, a cathode ray tube (CRT), or a printer for presenting text or images, anda pointing device 916, such as a mouse, a trackball, cursor direction keys, or a motion sensor, forcontrolling a position of a small cursor image presented on the display 914 and issuing commandsassociated with graphical elements presented on the display 914, and one or more camera sensors994 for capturing, recording and causing to store one or more still and / or moving images (e.g.,videos, movies, etc.) which also may comprise audio recordings. In some embodiments, forexample, in embodiments in which the computer system 900 performs all functions automaticallywithout human input, one or more of external input device 912, a display device 914 and pointingdevice 916 may be omitted.

[0158] In the illustrated embodiment, special purpose hardware, such as an application specificintegrated circuit (ASIC) 920, is coupled to bus 910. The special purpose hardware is configuredAttorney Docket No.: P10355PC00 Patentto perform operations not performed by processor 902 quickly enough for special purposes.Examples of ASICs include graphics accelerator cards for generating images for display 914,cryptographic boards for encrypting and decrypting messages sent over a network, speechrecognition, and interfaces to special external devices, such as robotic arms and medical scanningequipment that repeatedly perform some complex sequence of operations that are more efficientlyimplemented in hardware.

[0159] Computer system 900 also includes one or more instances of a communicationsinterface 970 coupled to bus 910. Communication interface 970 provides a one-way or two-waycommunication coupling to a variety of external devices that operate with their own processors,such as printers, scanners, and external disks. In general, the coupling is with a network link 978that is connected to a local network 980 to which a variety of external devices with their ownprocessors are connected. For example, communication interface 970 may be a parallel port or aserial port or a universal serial bus (USB) port on a personal computer. In some embodiments,communications interface 970 provides an information communication connection to acorresponding type of telephone line. In some embodiments, a communication interface 970 is acable modem that converts signals on bus 910 into signals for a communication connection over acoaxial cable or into optical signals for a communication connection over a fiber optic cable. Asanother example, communications interface 970 may be a local area network (LAN) card toprovide a data communication connection to a compatible LAN, such as Ethernet. Wireless linksmay also be implemented. For wireless links, the communications interface 970 sends or receivesor both sends and receives electrical, acoustic or electromagnetic signals, including infrared andoptical signals, that carry information streams, such as digital data. For example, in wirelesshandheld devices, such as mobile telephones like cell phones, the communications interface 970includes a radio band electromagnetic transmitter and receiver called a radio transceiver. In certainembodiments, the communications interface 970 enables connection to the communicationnetwork 97 in support of the AI-based decision support platform 101.

[0160] The term “computer-readable medium” as used herein refers to any medium thatparticipates in providing information to processor 902, including instructions for execution. Sucha medium may take many forms, including, but not limited to a computer-readable storage mediumAttorney Docket No.: P10355PC00 Patent(e.g., non-volatile media, volatile media), and transmission media. Non-transitory media, such asnon-volatile media, include, for example, optical or magnetic disks, such as storage device 908.Volatile media include, for example, dynamic memory 904. Transmission media include, forexample, twisted pair cables, coaxial cables, copper wire, fiber optic cables, and carrier waves thattravel through space without wires or cables, such as acoustic waves and electromagnetic waves,including radio, optical and infrared waves. Signals include man-made transient variations inamplitude, frequency, phase, polarization or other physical properties transmitted through thetransmission media. Common forms of computer-readable media include, for example, a floppydisk, a flexible disk, hard disk, magnetic tape, any other magnetic medium, a CD-ROM, CDRW,DVD, any other optical medium, punch cards, paper tape, optical mark sheets, any other physicalmedium with patterns of holes or other optically recognizable indicia, a RAM, a PROM, anEPROM, a FLASH-EPROM, an EEPROM, a flash memory, any other memory chip or cartridge,a carrier wave, or any other medium from which a computer can read. The term computer-readablestorage medium is used herein to refer to any computer-readable medium except transmissionmedia.

[0161] Logic encoded in one or more tangible media includes one or both of processorinstructions on a computer-readable storage media and special purpose hardware, such as ASIC920.

[0162] Network link 978 typically provides information communication using transmissionmedia through one or more networks to other devices that use or process the information. Forexample, network link 978 may provide a connection through local network 980 to a host computer982 or to equipment 984 operated by an Internet Service Provider (ISP). ISP equipment 984 inturn provides data communication services through the public, world-wide packet-switchingcommunication network of networks now commonly referred to as the Internet 990.

[0163] A computer called a server host 992 connected to the Internet hosts a process thatprovides a service in response to information received over the Internet. For example, server host992 hosts a process that provides information representing video data for presentation at displayAttorney Docket No.: P10355PC00 Patent914. It is contemplated that the components of system 900 can be deployed in variousconfigurations within other computer systems, e.g., host 982 and server 992.

[0164] At least some embodiments of the invention are related to the use of computer system900 for implementing some or all of the techniques described herein. According to oneembodiment of the invention, those techniques are performed by computer system 900 in responseto processor 902 executing one or more sequences of one or more processor instructions containedin memory 904. Such instructions, also called computer instructions, software and program code,may be read into memory 904 from another computer-readable medium such as storage device908 or network link 978. Execution of the sequences of instructions contained in memory 904causes processor 902 to perform one or more of the method steps described herein. In alternativeembodiments, hardware, such as ASIC 920, may be used in place of or in combination withsoftware to implement the invention. Thus, embodiments of the invention are not limited to anyspecific combination of hardware and software, unless otherwise explicitly stated herein.

[0165] The signals transmitted over network link 978 and other networks throughcommunications interface 970, carry information to and from computer system 900. Computersystem 900 can send and receive information, including program code, through the networks 980,990 among others, through network link 978 and communications interface 970. In an exampleusing the Internet 990, a server host 992 transmits program code for a particular application,requested by a message sent from computer 900, through Internet 990, ISP equipment 984, localnetwork 980 and communications interface 970. The received code may be executed by processor902 as it is received, or may be stored in memory 904 or in storage device 908 or any other non-volatile storage for later execution, or both. In this manner, computer system 900 may obtainapplication program code in the form of signals on a carrier wave.

[0166] Various forms of computer readable media may be involved in carrying one or moresequence of instructions or data or both to processor 902 for execution. For example, instructionsand data may initially be carried on a magnetic disk of a remote computer such as host 982. Theremote computer loads the instructions and data into its dynamic memory and sends theinstructions and data over a telephone line using a modem. A modem local to the computer systemAttorney Docket No.: P10355PC00 Patent900 receives the instructions and data on a telephone line and uses an infra-red transmitter toconvert the instructions and data to a signal on an infra-red carrier wave serving as the networklink 978. An infrared detector serving as communications interface 970 receives the instructionsand data carried in the infrared signal and places information representing the instructions and dataonto bus 910. Bus 910 carries the information to memory 904 from which processor 902 retrievesand executes the instructions using some of the data sent with the instructions. The instructionsand data received in memory 904 may optionally be stored on storage device 908, either before orafter execution by the processor 902.

[0167] FIG. 10 illustrates a chip set or chip 1000 upon which various embodiments of theinvention may be implemented. Chip set 1000 is programmed to the processes (e.g., FIG. 4) asdescribed herein and includes, for instance, the processor and memory components described withrespect to FIG. 9 incorporated in one or more physical packages (e.g., chips). By way of example,a physical package includes an arrangement of one or more materials, components, and / or wireson a structural assembly (e.g., a baseboard) to provide one or more characteristics such as physicalstrength, conservation of size, and / or limitation of electrical interaction. It is contemplated that incertain embodiments the chip set 1000 can be implemented in a single chip. It is furthercontemplated that in certain embodiments the chip set or chip 1000 can be implemented as a single“system on a chip.” It is further contemplated that in certain embodiments a separate ASIC wouldnot be used, for example, and that all relevant functions as disclosed herein would be performedby a processor or processors. Chip set or chip 1000, or a portion thereof, constitutes a means forperforming one or more steps of providing user interface navigation information associated withthe availability of functions. Chip set or chip 1000, or a portion thereof, constitutes a means forperforming one or more steps of providing decision support.

[0168] In one embodiment, the chip set or chip 1000 includes a communication mechanismsuch as a bus 1001 for passing information among the components of the chip set 1000. Aprocessor 1003 has connectivity to the bus 1001 to execute instructions and process informationstored in, for example, a memory 1005. The processor 1003 may include one or more processingcores with each core configured to perform independently. A multi-core processor enablesmultiprocessing within a single physical package. Examples of a multi-core processor includeAttorney Docket No.: P10355PC00 Patenttwo, four, eight, or greater numbers of processing cores. Alternatively or in addition, the processor1003 may include one or more microprocessors configured in tandem via the bus 1001 to enableindependent execution of instructions, pipelining, and multithreading. The processor 1003 mayalso be accompanied with one or more specialized components to perform certain processingfunctions and tasks such as one or more digital signal processors (DSP) 1007, or one or moreapplication-specific integrated circuits (ASIC) 1009. A DSP 1007 typically is configured toprocess real-world signals (e.g., sound) in real time independently of the processor 1003.Similarly, an ASIC 1009 can be configured to performed specialized functions not easilyperformed by a more general purpose processor. Other specialized components to aid inperforming the inventive functions described herein may include one or more field programmablegate arrays (FPGA), one or more controllers, or one or more other special-purpose computer chips.

[0169] In one embodiment, the chip set or chip 1000 includes merely one or more processorsand some software and / or firmware supporting and / or relating to and / or for the one or moreprocessors.

[0170] The processor 1003 and accompanying components have connectivity to the memory1005 via the bus 1001. The memory 1005 includes both dynamic memory (e.g., RAM, magneticdisk, writable optical disk, etc.) and static memory (e.g., ROM, CD-ROM, etc.) for storingexecutable instructions that when executed perform the inventive steps described herein to provideproviding decision support. The memory 1005 also stores the data associated with or generatedby the execution of the inventive steps.

[0171] FIG. 11 is a diagram of exemplary components of a mobile terminal (e.g., handset) forcommunications, which is capable of operating in the system of FIG. 1, according to oneembodiment. In some embodiments, mobile terminal 1101, or a portion thereof, constitutes ameans for performing one or more steps of the described processes. Generally, a radio receiver isoften defined in terms of front-end and back-end characteristics. The front-end of the receiverencompasses all of the Radio Frequency (RF) circuitry whereas the back-end encompasses all ofthe base-band processing circuitry. As used in this application, the term “circuitry” refers to both:(1) hardware-only implementations (such as implementations in only analog and / or digitalAttorney Docket No.: P10355PC00 Patentcircuitry), and (2) to combinations of circuitry and software (and / or firmware) (such as, ifapplicable to the particular context, to a combination of processor(s), including digital signalprocessor(s), software, and memory(ies) that work together to cause an apparatus, such as a mobilephone or server, to perform various functions). This definition of “circuitry” applies to all uses ofthis term in this application, including in any claims. As a further example, as used in thisapplication and if applicable to the particular context, the term “circuitry” would also cover animplementation of merely a processor (or multiple processors) and its (or their) accompanyingsoftware / or firmware. The term “circuitry” would also cover if applicable to the particular context,for example, a baseband integrated circuit or applications processor integrated circuit in a mobilephone or a similar integrated circuit in a cellular network device or other network devices.

[0172] Pertinent internal components of the telephone include a Main Control Unit (MCU)1103, a Digital Signal Processor (DSP) 1105, and a receiver / transmitter unit including amicrophone gain control unit and a speaker gain control unit. A main display unit 1107 providesa display to the user in support of various applications and mobile terminal functions that performor support the steps of providing decision support. The display 1107 includes display circuitryconfigured to display at least a portion of a user interface of the mobile terminal (e.g., mobiletelephone). Additionally, the display 1107 and display circuitry are configured to facilitate usercontrol of at least some functions of the mobile terminal. An audio function circuitry 1109 includesa microphone 1111 and microphone amplifier that amplifies the speech signal output from themicrophone 1111. The amplified speech signal output from the microphone 1111 is fed to acoder / decoder (CODEC) 1113.

[0173] A radio section 1115 amplifies the power and converts frequency in order tocommunicate with a base station, which is included in a mobile communication system, viaantenna 1117. The power amplifier (PA) 1119 and the transmitter / modulation circuitry areoperationally responsive to the MCU 1103, with an output from the PA 1119 coupled to theduplexer 1121 or circulator or antenna switch, as known in the art. The PA 1119 also couples toa battery interface and power control unit 1120.Attorney Docket No.: P10355PC00 Patent

[0174] In use, a user of mobile terminal 1101 speaks into the microphone 1111 and his or hervoice along with any detected background noise is converted into an analog voltage. The analogvoltage is then converted into a digital signal through the Analog to Digital Converter (ADC) 1123.The control unit 1103 routes the digital signal into the DSP 1105 for processing therein, such asspeech encoding, channel encoding, encrypting, and interleaving. In one embodiment, theprocessed voice signals are encoded, by units not separately shown, using a cellular transmissionprotocol such as enhanced data rates for global evolution (EDGE), general packet radio service(GPRS), global system for mobile communications (GSM), Internet protocol multimediasubsystem (IMS), universal mobile telecommunications system (UMTS), etc., as well as any othersuitable wireless medium, e.g., microwave access (WiMAX), Long Term Evolution (LTE)networks, code division multiple access (CDMA), wideband code division multiple access(WCDMA), wireless fidelity (WiFi), satellite, and the like, or any combination thereof.

[0175] The encoded signals are then routed to an equalizer 1125 for compensation of anyfrequency-dependent impairments that occur during transmission though the air such as phase andamplitude distortion. After equalizing the bit stream, the modulator 1127 combines the signal withan RF signal generated in the RF interface 1129. The modulator 1127 generates a sine wave byway of frequency or phase modulation. In order to prepare the signal for transmission, an up-converter 1131 combines the sine wave output from the modulator 1127 with another sine wavegenerated by a synthesizer 1133 to achieve the desired frequency of transmission. The signal isthen sent through a PA 1119 to increase the signal to an appropriate power level. In practicalsystems, the PA 1119 acts as a variable gain amplifier whose gain is controlled by the DSP 1105from information received from a network base station. The signal is then filtered within theduplexer 1121 and optionally sent to an antenna coupler 1135 to match impedances to providemaximum power transfer. Finally, the signal is transmitted via antenna 1117 to a local base station.An automatic gain control (AGC) can be supplied to control the gain of the final stages of thereceiver. The signals may be forwarded from there to a remote telephone which may be anothercellular telephone, any other mobile phone or a land-line connected to a Public SwitchedTelephone Network (PSTN), or other telephony networks.Attorney Docket No.: P10355PC00 Patent

[0176] Voice signals transmitted to the mobile terminal 1101 are received via antenna 1117and immediately amplified by a low noise amplifier (LNA) 1137. A down-converter 1139 lowersthe carrier frequency while the demodulator 1141 strips away the RF leaving only a digital bitstream. The signal then goes through the equalizer 1125 and is processed by the DSP 1105. ADigital to Analog Converter (DAC) 1143 converts the signal and the resulting output is transmittedto the user through the speaker 1145, all under control of a Main Control Unit (MCU) 1103 whichcan be implemented as a Central Processing Unit (CPU).

[0177] The MCU 1103 receives various signals including input signals from the keyboard1147. The keyboard 1147 and / or the MCU 1103 in combination with other user input components(e.g., the microphone 1111) comprise a user interface circuitry for managing user input. The MCU1103 runs a user interface software to facilitate user control of at least some functions of the mobileterminal 1101 to provide decision support. The MCU 1103 also delivers a display command anda switch command to the display 1107 and to the speech output switching controller, respectively.Further, the MCU 1103 exchanges information with the DSP 1105 and can access an optionallyincorporated SIM card 1149 and a memory 1151. In addition, the MCU 1103 executes variouscontrol functions required of the terminal. The DSP 1105 may, depending upon theimplementation, perform any of a variety of conventional digital processing functions on the voicesignals. Additionally, DSP 1105 determines the background noise level of the local environmentfrom the signals detected by microphone 1111 and sets the gain of microphone 1111 to a levelselected to compensate for the natural tendency of the user of the mobile terminal 1101.

[0178] The CODEC 1113 includes the ADC 1123 and DAC 1143. The memory 1151 storesvarious data including call incoming tone data and is capable of storing other data including musicdata received via, e.g., the global Internet. The software module could reside in RAM memory,flash memory, registers, or any other form of writable storage medium known in the art. Thememory device 1151 may be, but not limited to, a single memory, CD, DVD, ROM, RAM,EEPROM, optical storage, magnetic disk storage, flash memory storage, or any other non-volatilestorage medium capable of storing digital data.Attorney Docket No.: P10355PC00 Patent

[0179] An optionally incorporated SIM card 1149 carries, for instance, important information,such as the cellular phone number, the carrier supplying service, subscription details, and securityinformation. The SIM card 1149 serves primarily to identify the mobile terminal 1101 on a radionetwork. The card 1149 also contains a memory for storing a personal telephone number registry,text messages, and user specific mobile terminal settings.

[0180] Further, one or more camera sensors 1153 may be incorporated onto the mobile station1101 wherein the one or more camera sensors may be placed at one or more locations on the mobilestation. Generally, the camera sensors may be utilized to capture, record, and cause to store oneor more still and / or moving images (e.g., videos, movies, etc.) which also may comprise audiorecordings.

[0181] While the invention has been described in connection with a number of embodimentsand implementations, the invention is not so limited but covers various obvious modifications andequivalent arrangements, which fall within the purview of the appended claims. Although featuresof the invention are expressed in certain combinations among the claims, it is contemplated thatthese features can be arranged in any combination and order.

Claims

Attorney Docket No.: P10355PC00 PatentCLAIMSWHAT IS CLAIMED IS:

1. A method comprising:receiving historical performance data and economic data associated with one or more illiquidassets; training a machine learning model using the historical performance data to predict one ormore features including cumulative contributions, distributions, and net asset value(NAV) percentage changes;incorporating regime-based transition probabilities into the machine learning model toaccount for different economic stages; andgenerating cash flow forecasts and stress testing output for the illiquid assets based on thetrained model.

2. The method of claim 1, wherein the economic data includes macroeconomic indicators, themethod further comprising:classifying the macroeconomic indicators to represent one or more economic sectors as areduced set of macroeconomic indicators.

3. The method of claim 1, further comprising:generating a covariance matrix for the reduced set of macroeconomic indicators to determineexpected changes.

4. The method of claim 1, further comprising:generating a composite index using the historical performance data, wherein the compositeindex specifies contributions of a plurality of performance measurements for the one oror more illiquid assets.Attorney Docket No.: P10355PC00 Patent5. The method of claim 4, further comprising:generating another composite index to specify variability of the composite index, wherein theperformance of the one or more illiquid assets are represented according to the compositeindices.

6. The method of claim 1, wherein the generating of the cash flow forecast comprises:applying a best-fit deterministic model to the historical performance data and economic datato produce a first predicted output;modifying the best-fit deterministic model based on actual cash flow data to yield an errorterm as a second predicted output; andcombining the first predicted output and second predicted output as a final predicted output.

7. The method of claim 1, further comprising:computing slope for the one or more features, wherein the slope is utilized to determine thecash flow forecasts.

8. A system comprising:a memory configured to store computer-executable instructions; andone or more processors configured to execute the instructions to:receive historical performance data and economic data associated with one or moreilliquid assets;train a machine learning model using the historical performance data to predict one ormore features including cumulative contributions, distributions, and net asset value(NAV) percentage changes;incorporate regime-based transition probabilities into the machine learning model toaccount for different economic stages; andgenerate cash flow forecasts and stress testing output for the illiquid assets based on thetrained model.Attorney Docket No.: P10355PC00 Patent9. The system of claim 8, wherein the economic data includes macroeconomic indicators,wherein the one or more processors are further configured to execute the instructions to:classify the macroeconomic indicators to represent one or more economic sectors as areduced set of macroeconomic indicators.

10. The system of claim 8, wherein the one or more processors are further configured toexecute the instructions to:generate a covariance matrix for the reduced set of macroeconomic indicators to determineexpected changes.

11. The system of claim 8, wherein the one or more processors are further configured toexecute the instructions to:generate a composite index using the historical performance data, wherein the compositeindex specifies contributions of a plurality of performance measurements for the one oror more illiquid assets.

12. The system of claim 11, wherein the one or more processors are further configured toexecute the instructions to:generate another composite index to specify variability of the composite index, wherein theperformance of the one or more illiquid assets are represented according to the compositeindices.

13. The system of claim 8, wherein the generation of the cash flow forecast comprises:applying a best-fit deterministic model to the historical performance data and economic datato produce a first predicted output;modifying the best-fit deterministic model based on actual cash flow data to yield an errorterm as a second predicted output; andcombining the first predicted output and second predicted output as a final predicted output.Attorney Docket No.: P10355PC00 Patent14. The system of claim 8, wherein the one or more processors are further configured toexecute the instructions to:compute slope for the one or more features, wherein the slope is utilized to determine thecash flow forecasts.

15. An apparatus comprising:at least one processor; andat least one memory including computer program code for one or more programs,the at least one memory and the computer program code configured to, with the at leastone processor, cause the apparatus to perform at least the following,receive historicalperformance data and economic data associated with one or more illiquid assets;train a machine learning model using the historical performance data to predict one ormore features including cumulative contributions, distributions, and net asset value(NAV) percentage changes;incorporate regime-based transition probabilities into the machine learning model toaccount for different economic stages; andgenerate cash flow forecasts and stress testing output for the illiquid assets based on thetrained model.

16. The apparatus of claim 15, wherein the economic data includes macroeconomicindicators, wherein the one or more processors are further configured to execute the instructionsto: classify the macroeconomic indicators to represent one or more economic sectors as areduced set of macroeconomic indicators.

17. The apparatus of claim 15, wherein the one or more processors are further configured toexecute the instructions to:generate a covariance matrix for the reduced set of macroeconomic indicators to determineexpected changes; andAttorney Docket No.: P10355PC00 Patentcompute slope for the one or more features, wherein the slope is utilized to determine thecash flow forecasts.

18. The apparatus of claim 15, wherein the one or more processors are further configured toexecute the instructions to:generate a composite index using the historical performance data, wherein the compositeindex specifies contributions of a plurality of performance measurements for the one oror more illiquid assets.

19. The apparatus of claim 11, wherein the one or more processors are further configured toexecute the instructions to:generate another composite index to specify variability of the composite index, wherein theperformance of the one or more illiquid assets are represented according to the compositeindices.

20. The apparatus of claim 15, wherein the generation of the cash flow forecast comprises:applying a best-fit deterministic model to the historical performance data and economic datato produce a first predicted output;modifying the best-fit deterministic model based on actual cash flow data to yield an errorterm as a second predicted output; andcombining the first predicted output and second predicted output as a final predicted output.

Images