Method and system for predicting a value

Abstract

A system has a trained database of one or more trained identified values which is controlled by a processor by applying machine learning techniques on the one or more identified values, the trained database providing at least one trained identified value along with at least one variable value of one or more different segments to a machine learning model. The system has a machine learning engine processing the trained identified value with at least one variable value for generating an estimated value for each variable segment. The system has a processing unit processing the estimated values of each variable segment along with the trained identified values of that variable segment to create plurality of variable segment specific contributions wherein the processing unit is generating a forecast value of at least one variable for a target period by applying specific machine learning techniques on at least one segment specific contributions.

Description

METHOD AND SYSTEM FOR PREDICTING A VALUE

[0001] The present invention relates to a method and system for predicting a value based on a query made by a user. More particularly, the present invention relates to a method and system of predicting a value based on processing of one or more identified values.

[0002] With the evolution of technical elements such as technical processes various conventional processes are simplified. The advancement in the field of machine learning and artificial intelligence changed the landscape of various conventional techniques and enabled complex processes to be solved. Data processing is one of the processes which is widely used by using machine learning and artificial intelligence. The data processing is used in many aspects and one of the aspects is to generate a forecast value or to predict a value based on the historical data.

[0003] There are many conventional techniques available for predicting a data value for a specific period of time. However, the conventional techniques are not able to provide accurate results and requires lot of time and resources for processing a data set and the processing become more difficult if the dataset is voluminous. The generation of an accurate result from a voluminous dataset is hardly achievable and therefore requires lot of resources. Additionally, to predict or to forecast a value specific to a cashflow depending on historical data for a future period requires lot of precision processing.

[0004] There are very few research papers available which discloses considering few factors in order to improve an accuracy in order to forecast a value. However, there are hardly any machine learning based models which emphasizes on accuracy by processing the historical dataset. Most models have mentioned about R-square of models which is basically the power of independent variable to explain the variability of the target variable. Higher R-square does not always lead to higher accuracy.

[0005] The existing papers have mostly modelled operating cash flow using all its components as independent variable or have replaced some of the components with variables like sales or operating profit. Research has also found that disaggregating cash flows into its major components does not appear to enhance cash flow prediction. Models incorporating income statement information seem to have poor out-of-sample / test sample predictions.

[0006] One of factors identified for not having accurate results is the due to not considering of delay (in collections or payments) metrics which leads to significant error in model accuracy. Most existing models are built using ARIMA with exogenous variables (ARIMAX model) or with multiple regression models or with multivariate time-series regression model (MULT).

[0007] Therefore, there is a need to provide a technical method and system for predicting a value considering the variable factors which varies with a period of time, and which are not processed by the existing models and systems.

[0008] The following presents a simplified summary of the subject matter in order to provide a basic understanding of some aspects of subject matter embodiments. This summary is not an extensive overview of the subject matter. It is not intended to identify key / critical elements of the embodiments or to delineate the scope of the subject matter. Its sole purpose to present some concepts of the subject matter in a simplified form as a prelude to the more detailed description that is presented later.

[0009] The object of the present invention is to provide a method and system for predicting a value based on processing of one or more identified values.

[0010] The present invention is comprises of a trained database, one or more processors, a machine learning engine, a processing unit, and an analytical module.

[0011] In an embodiment of the present invention, the method of the present invention is predicting a value based on processing of one or more identified values. The method applies, through a processor, machine learning techniques on the one or more identified values for generating a trained database of one or more trained identified values. The method further provides, by the trained database, at least one trained identified value along with at least one variable value of one or more different segments to a machine learning model. The method further processes the trained identified value with at least one variable value by a machine learning engine for generating an estimated value for each variable segment. The method further processes, by a processing unit, the estimated values of each variable segment along with the trained identified values of that variable segment to create plurality of variable segment specific contributions wherein the processing unit is generating a forecast value of at least one variable for a target period by applying specific machine learning techniques on at least one segment specific contributions.

[0012] In an embodiment of the present invention, the one or more identified values are the numerical historical data values of a cashflow component and are provided by a user and the at least one variable of one or more segments is a numerical data value of a present period and is comprises of different metrics including a delay metric and cashflow components.

[0013] In another embodiment of the present invention, the machine learning engine is being trained by processing the at least one variable value of one or more different segments with the trained identified values based on a trend and / or pattern information of the at least one variable value of one or more different segments, the trend and pattern information being dynamically updated by the user.

[0014] Another embodiment of the present invention analysing, by an analytical module of the processing unit, the effect of trend and pattern information of the at least one variable value of one or more different segments on the historical data values of a cashflow component wherein the plurality of variable segment specific contributions is updated for each trend and pattern information based on change of the different metrics.

[0015] Yet another embodiment of the present invention processing, by the processing unit, the updated segment specific contributions to update the forecast value in the event of change in the trend and pattern information of the at least one variable value of one or more different segments.

[0016] The present invention in an embodiment also discloses a system for predicting a value based on processing of one or more identified values. The system is comprised of a trained database of one or more trained identified values which is controlled by a processor by applying machine learning techniques on the one or more identified values. The trained database is configured to provide at least one trained identified value along with at least one variable value of one or more different segments to a machine learning model. Further, a machine learning engine, processing the trained identified value with at least one variable value for generating an estimated value for each variable segment. A processing unit, processing the estimated values of each variable segment along with the trained identified values of that variable segment to create plurality of variable segment specific contributions wherein the processing unit, generating a forecast value of at least one variable for a target period by applying specific machine learning techniques on at least one segment specific contributions.

[0017] The foregoing and further objects, features and advantages of the present subject matter will become apparent from the following description of exemplary embodiments with reference to the accompanying drawings, wherein like numerals are used to represent like elements.

[0018] It is to be noted, however, that the appended drawings along with the reference numerals illustrate only typical embodiments of the present subject matter, and are therefore, not to be considered for limiting of its scope, for the subject matter may admit to other equally effective embodiments.

[0019] Fig. 1 shows a flow diagram of the different steps involved in predicting a value based on processing of one or more identified values, according to an exemplary embodiment.Fig. 2 shows a schematic diagram that shows the different modules predicting a value based on processing of one or more identified values, according to an exemplary embodiment.

[0020] The present invention is directed to a method and system for predicting a value based on processing of one or more identified values. The system is comprised of a trained database of one or more trained identified values which is controlled by a processor by applying machine learning techniques on the one or more identified values. The trained database is configured to provide at least one trained identified value along with at least one variable value of one or more different segments to a machine learning model. Further, a machine learning engine, processing the trained identified value with at least one variable value for generating an estimated value for each variable segment. A processing unit, processing the estimated values of each variable segment along with the trained identified values of that variable segment to create plurality of variable segment specific contributions wherein the processing unit, generating a forecast value of at least one variable for a target period by applying specific machine learning techniques on at least one segment specific contributions.

[0021] Fig. 1 shows a flow diagram of the different steps involved in predicting a value based on processing of one or more identified values according to an exemplary embodiment. Specifically, Fig. 1 shows a method 100 for predicting a value based on processing of one or more identified values. As shown Fig. 1, the method 100 starts at step 101, by applying machine learning techniques on the one or more identified values for generating a trained database of one or more trained identified values. A person skilled in the art should appreciate that the machine learning techniques can be applied with a processor. The processor is enabled to process the multiple data of structured and of unstructured form. Here, in the present invention a trained database is referred as the historical data which comprises of historical cash flows and / or data relevant to the interest of the user. For cashflow, generally the data is monthly or quarterly. In cashflow data, medium to large fluctuations are normal and hence, defining changepoints is a key thing to forecast cashflows. Therefore, the present invention having a machine learning based cashflow prediction using the OSS based explainable AI or a PROPHET model as it has multiple hyperparameters and many options to include different sort of information into the model like changepoints which are points in the series where fluctuations in the data is more than the average fluctuation. At step 102, the present invention is enabled to provide by the trained database at least one trained identified value along with at least one variable value of one or more different segments to a machine learning model. Here, the machine learning model is realized, for example, by a neural network. The neural network includes a plurality of artificial neurons and has a plurality of synapses that respectively connect the plurality of artificial neurons. Each of the plurality of synapses has a weight. When the neural network receives an input, it performs calculations using the weights associated with each of the plurality of synapses and outputs according to the input. A model using the neural network that represents the connection relationship between the neurons and the synapses is stored in a memory, for example, in the form of software. The model may also be realized as a dedicated circuit. Similarly, the weights of each synapse are also stored in a memory in the form of software. In addition, a circuit representing the weights may be further implemented in the dedicated circuit. Note that, when a machine learning model is configured using a plurality of models, it is not necessarily necessary that all the models are stored in the same memory. There are various models that use the neural network. A wide variety of models, such as a Transformer, a convolutional neural network (CNN), and a recurrent neural network (RNN), may be adopted or replaced to realize the machine learning model. In addition, the machine learning engine may be configured integrally with the above-mentioned machine learning model, or may be implemented separately. The one or more variable value could be cashflow data of present month. It could be delay data, PMI data and even combinations of these data. Since, these data varies for every month therefore these are named as variable data which is varying with respect to different time periods. It is to be understood that the delay data is at least one of: % of delayed collections out of overall organisation collections estimate in a particular month and / or % of projects delayed out of overall organisation projects in a particular month. The Purchasing Managers' Index (PMI) is a weighted average of the following five indices: New Orders (30%), Output (25%), Employment (20%), Suppliers' Delivery Times (15%) and Stocks of Purchases (10%).

[0022] The ML model of the present invention processes the historical data and the variable data wherein the time series technique are applied on these data processing. At step 103, the present invention processing the trained identified value with at least one variable value by a machine learning engine for generating an estimated value for each variable segment. A person skilled in the art would appreciate that this step of the present invention is processing the output of the ML model wherein the trained identified value is processed with a variable value. The variable value could be seasonality and trend information. Seasonality is estimated using a partial Fourier sum. The number of terms in the partial sum (the order) is a parameter that determines how quickly the seasonality can change. This order can be tuned as hyperparameter to improve the accuracy. Prior scales can be used to reduce overfitting or underfitting in trend and seasonality factors. This can also be tuned as hyperparameters. Further, it is described herein that the native decomposition of the forecast into trend, seasonalities, events and extra-regressors components is meaningful for low-tech profiles. Prophet is an additive model wherein the sum of each component equals the prediction and it provides both local and global explainability. Further, the methods like SHAP or Shapley Values may be used to explain the contribution of each component. Further, for an improved accuracy, the monthly data of minimum two years to cover the yearly seasonality (if present) and maximum of 4 years should be taken. Quarterly data if taken would require around 24 data points to build a stable and robust model which translates to minimum of 6 years of data. The ML engine is processes the trained identified value with the one variable in way that it is able to provide and link one identified value with multiple variables or multiple identified values with one variable so that all the permutations and combinations are worked out to cover the aspect of the data. At step 104, another ML engine processes the estimated values of each variable segment along with the trained identified values of that variable segment to create plurality of variable segment specific contributions. The processing of the estimated value which is generated on the processing of the trained identified value with one variable value of each segment enables to provide more detailed analysis by which a user can expect to have more accurate and reliable results as per the user need. At step 105 a processing unit generates a forecast value of at least one variable for a target period by applying specific machine learning techniques on at least one segment specific contributions. The target period is set by the user. For example: the individual estimate of each segment is then added together to create the forecast as desired by the user. The machine learning techniques applied by the machine learning engine is in communication with a processor and an analytical module wherein the analytical module analysing the effect of trend and pattern information of the at least one variable value of one or more different segments on the historical data values of a cashflow component. The analytical module analyzation is based on the changes of the data inputted and even the data as managed in the historical database. The analytical module is also able to analyse the macro-economic factors such as Purchasing managaers' index (PMI) data, Services Business activity Index data, Total Activity Index data, Business Expectation (BEI) data and / or Government spending data.

[0023] The Purchasing Managers' Index (PMI) is a weighted average of the following five indices: New Orders (30%), Output (25%), Employment (20%), Suppliers' Delivery Times (15%) and Stocks of Purchases (10%). Services Business Activity Index for services sector is the index which measures the changes in the volume of business activity compared with previous month. Total Activity Index for construction sector is the index which measures the changes in the total volume of construction activity compared with one month previously. Business Expectation Index (BEI) gives a snapshot of the business outlook in every quarter and takes values between 0 and 200, with 100 being the threshold separating expansion from contraction. Business Expectations Index (BEI) is calculated as a weighted net response of nine business indicators. The nine indicators considered are: (1) overall business situation; (2) production; (3) order books; (4) inventory of raw material; (5) inventory of finished goods; (6) profit margins; (7) employment; (8) exports; and (9) capacity utilisation. BEI gives a snapshot of the business outlook in every quarter and takes values between 0 and 200, with 100 being the threshold separating expansion from contraction. Government spending is another indicator which can be used to measure cashflow especially if the organisation works on government projects. More expenditure by government might lead to more projects for organisation which might influence cashflow. It is to be noted that by incorporating the volatile macro-economic situation which is one of the best indicators to be included in the model is providing an accurate and timely insight into the health of the global economy. Further, generating a cashflow forecast model using collections hold back data and economic volatility data of that country is also dependent on the deferred / postponed / held back collections which is incorporated through two metrics:

[0024] a) % of deferred / postponed / held back collections out of overall organisation collections estimate in a particular month;

[0025] b) % of projects deferred / postponed / held back out of overall organisation projects in a particular month.

[0026] The economic volatility of a country is included in the model using monthly Purchasing Managers' Index (PMI) data, Services Business Activity Index data, Total Activity Index data, Business Expectation Index (BEI) data and Government spending data.

[0027] The present invention is using fewer cashflow components viz. Collections, Cost of goods sold (COGS) and Selling, General and Administrative expenses (SGA) and then use the same metrics from different business units to improve the overall accuracy. Further, transformations may be applied on these variables to improve the overall accuracy. Also, the change variables derived from these variables may also be used as additional regressors to the model. Further for the better understanding on the delay data as well as the historical data associated with the delay data the delay information is included as separate variables would be in the form of % of delayed collections out of overall organisation collections estimate in a particular month and / or % of projects delayed out of overall organisation projects in a particular month. Further, the delay information may be calculated by aggregating project level information at BU level where delay information is available. This information is then incorporated in the model to learn from historic patterns to forecast the future expected delays more accurately.

[0028] In order to provide more accurate forecasts, the macro-economic factors are considered wherein an index provided by S&P Global may be used. The index takes multiple macro-economic factors to create a single metric. The indices vary between 0 and 100, with a reading above 50 indicating an overall increase compared to the previous month, and below 50 an overall decrease. It also has Services Business Activity Index for services sector which measures the changes in the volume of business activity compared with previous month. In addition, there is Total Activity Index for construction sector which measures the changes in the total volume of construction activity compared with one month previously.

[0029] In another embodiment, Fig. 2 shows different modules of a system 200, predicting a value based on processing of one or more identified values, according to an exemplary embodiment. The different modules are trained database 202, one or more processors 204, a ML model 206, a machine learning engine 208, a processing unit 210, and an analytical module 212.

[0030] The system 200 for predicting a value based on processing of one or more identified values is having a trained database 202. The trained database is controlled by a processor, not shown in figure, and comprises of one or more trained identified values on which machine learning techniques are applied with the help of the processors on the one or more identified values. The processor of the trained database is also configured to process and provide the trained identified values along with at least one variable value of one or more different segments to a machine learning model 206. As already discussed, that a trained database is referred as the historical data which comprises of historical cash flows and / or data relevant to the interest of the user. For cashflow, generally the data is monthly or quarterly. In cashflow data, medium to large fluctuations are normal and hence, defining changepoints is a key thing to forecast cashflows. Therefore, the present invention having a machine learning based cashflow prediction using the OSS based explainable AI or a PROPHET model as it has multiple hyperparameters and many options to include different sort of information into the model like changepoints which are points in the series where fluctuations in the data is more than the average fluctuation. The processor is also able to process and store the data in the format as desired by a user. Further, the ML model of the present invention processes with the help of processor, the historical data and the variable data wherein the time series technique are applied on these data processing.

[0031] The machine learning engine 208 processing the trained identified value with at least one variable value for generating an estimated value for each variable segment. The output of the ML model wherein the trained identified value is processed with a variable value. The variable value could be seasonality and trend information. Seasonality is estimated using a partial Fourier sum. The number of terms in the partial sum (the order) is a parameter that determines how quickly the seasonality can change. This order can be tuned as hyperparameter to improve the accuracy. The ML engine 208 is processing the trained identified value with the one variable in way that it is able to provide and link one identified value with multiple variables or multiple identified values with one variable so that all the permutations and combinations are worked out to cover the aspect of the data. Further, the processing unit in conjunction with the ML engine processes the estimated values of each variable segment along with the trained identified values of that variable segment to create plurality of variable segment specific contributions. The processing of the estimated value which is generated on the processing of the trained identified value with one variable value of each segment enables to provide more detailed analysis by which a user can expect to have more accurate and reliable results as per the user need.

[0032] The processing unit 210 processing the estimated values of each variable segment along with the trained identified values of that variable segment to create plurality of variable segment specific contributions. The processing unit is also generating a forecast value of at least one variable for a target period by applying specific machine learning techniques on at least one segment specific contributions. The processing unit generates a forecast value of at least one variable for a target period by applying specific machine learning techniques on at least one segment specific contributions. The target period is set by the user. For example: the individual estimate of each segment is then added together to create the forecast as desired by the user. The machine learning techniques applied by the machine learning engine is in communication with a processor and an analytical module wherein the analytical module analysing the effect of trend and pattern information of the at least one variable value of one or more different segments on the historical data values of a cashflow component. The analytical module analyzation is based on the changes of the data inputted and even the data as managed in the historical database. The analytical module is also able to analyse the macro-economic factors such as Purchasing managaers' index (PMI) data, Services Business activity Index data, Total Activity Index data, Business Expectation (BEI) data and / or Government spending data. The Purchasing Managers' Index (PMI) is a weighted average of the following five indices: New Orders (30%), Output (25%), Employment (20%), Suppliers' Delivery Times (15%) and Stocks of Purchases (10%). Services Business Activity Index for services sector is the index which measures the changes in the volume of business activity compared with previous month.

[0033] Here, for the sake of clarification the one or more identified values are the numerical historical data values of a cashflow component and are provided by a user. Further, the variable of one or more segments is a numerical data value of a present period and is comprises of different metrics including a delay metric, cashflow components and macro-economic factors. All these have been sufficiently described in above paragraphs and are not repeated here. The person skilled in the art would appreciate that the machine learning engine is being trained by processing the at least one variable value of one or more different segments with the trained identified values based on a trend and / or pattern information of the at least one variable value of one or more different segments, the trend and pattern information being dynamically updated by the user. The pattern information is defined in the earlier paragraphs wherein the aspect of seasonality is also discussed.

[0034] The analytical module 212 of the processing unit is a key element in the present invention hardware enablement wherein the analytical module is analysing the effect of trend and pattern information of the at least one variable value of one or more different segments on the historical data values of a cashflow component, wherein the plurality of variable segment specific contributions is being updated for each trend and pattern information based on change of the different metrics. The machine learning techniques applied by the machine learning engine is in communication with a processor and an analytical module wherein the analytical module analysing the effect of trend and pattern information of the at least one variable value of one or more different segments on the historical data values of a cashflow component. The analytical module analyzation is based on the changes of the data inputted and even the data as managed in the historical database. The analytical module is also able to analyse the macro-economic factors such as Purchasing managaers' index (PMI) data, Services Business activity Index data, Total Activity Index data, Business Expectation (BEI) data and / or Government spending data. The Purchasing Managers' Index (PMI) is a weighted average of the following five indices: New Orders (30%), Output (25%), Employment (20%), Suppliers' Delivery Times (15%) and Stocks of Purchases (10%). Services Business Activity Index for services sector is the index which measures the changes in the volume of business activity compared with previous month.

[0035] Therefore, with the present invention it is possible to reduce the number of cashflow components as only few components are used to determine the forecast. Therefore, storage and processing of large data set is avoided and hence the present invention also able to provide an efficient, less-time consuming solution to predict / forecast the value for a target period as desired by the user. Further, by considering fewer cashflow components viz. Collections, Cost of goods sold (COGS) and Selling, General and Administrative expenses (SGA) and then use the same metrics from different business units improves the overall accuracy. Further, the present invention is also able to transform to improve the overall accuracy. Also, the change variables derived from these variables can also be used as additional regressors to the model.

[0036] When performing output in each embodiment (values obtained in each process such as the forecast value or the trained identified value), the display contents may be changed based on the display information of the output destination. The display information is, for example, the size of the display and the ratio of the vertical length to the horizontal length. Based on the display information, the display contents may be changed so that, for example, the larger the display size, the larger the size of the characters and figures such as graphs. At this time, an upper limit value may be set so that the display contents are not displayed larger than a predetermined size. On the other hand, the display contents may be changed so that the smaller the display size, the smaller the size of the characters and figures such as graphs. In this case, a lower limit value may be set so that the display contents are not displayed below a predetermined size. In addition, the display position within the screen on the display may be changed, or a specific item may not be displayed on the same screen.

[0037] As another aspect, the display contents may be changed according to the processing capacity of the information processing device that performs the processing for displaying on the display. For example, when the processing capacity is low, the displayed contents and the amount of information when displayed may be reduced compared to when the processing capacity is high. Regarding the processing capacity, a predetermined specification such as the memory size may be referenced, or the operating status of the processor and the execution status of the task may be referenced.

[0038] In addition, as another aspect, the forecast value may be output in the form of text, or may be output in the form of a diagram plotted on a graph, etc. The forecast value may also be output as a voice, etc.

[0039] As will be appreciated by one of skill in the art, the present invention may be embodied as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, a software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product on a computer-usable storage medium having computer-usable program code embodied in the medium.

[0040] Furthermore, the present invention was described in part above with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention.

[0041] It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions / acts specified in the flowchart and / or block diagram block or blocks.

[0042] These computer program instructions may also be stored in a computer- readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function / act specified in the flowchart and / or block diagram block or blocks.

[0043] The computer program instructions may also be loaded onto a computer or other programmable data processing apparatus like a scanner / check scanner to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions / acts specified in the flowchart and / or block diagram block or blocks.

[0044] The flowchart and schematic diagrams illustrate the architecture, functionality, and operations of some embodiments of methods, systems, and computer program products for managing security associations over a communication network. In this regard, each block may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that in other implementations, the function(s) noted in the blocks may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently or the blocks may sometimes be executed in the reverse order, depending on the functionality involved.

[0045] In the drawings and specification, there have been disclosed exemplary embodiments of the invention. Although specific terms are employed, they are used in a generic and descriptive sense only and not for purposes of limitation, the scope of the invention being defined by the following claims.

[0046] For example, the whole or part of the exemplary example embodiments disclosed above can be described as, but not limited to, the following supplementary notes. (Supplementary note 1) 　　A method of predicting a value based on processing of one or more identified values, comprising: 　　applying, by a processor, machine learning techniques on the one or more identified values for generating a trained database of one or more trained identified values; 　　providing, by the trained database, at least one trained identified value along with at least one variable value of one or more different segments to a machine learning model; 　　processing the trained identified value with at least one variable value by a machine learning engine for generating an estimated value for each variable segment; 　　processing, by a processing unit, the estimated values of each variable segment along with the trained identified values of that variable segment to create plurality of variable segment specific contributions; 　　generating, by the processing unit, a forecast value of at least one variable for a target period by applying specific machine learning techniques on at least one segment specific contributions. (Supplementary note 2) 　　The method according to supplementary note 1, wherein the one or more identified values are the numerical historical data values of a cashflow component and are provided by a user. (Supplementary note 3) 　　The method according to supplementary note 1, wherein the at least one variable of one or more segments is a numerical data value of a present period and is comprises of different metrics including a delay metric and cashflow components. (Supplementary note 4) 　　The method according to supplementary note 3, wherein the machine learning engine is being trained by processing the at least one variable value of one or more different segments with the trained identified values based on a trend and / or pattern information of the at least one variable value of one or more different segments, the trend and pattern information being dynamically updated by the user. (Supplementary note 5) 　　The method according to supplementary note 4, comprises: 　　Analysing, by an analytical module of the processing unit, the effect of trend and pattern information of the at least one variable value of one or more different segments on the historical data values of a cashflow component. (Supplementary note 6) 　　The method according to supplementary note 5, wherein the plurality of variable segment specific contributions is updated for each trend and pattern information based on change of the different metrics. (Supplementary note 7) 　　The method according to supplementary note 1, wherein the target period is set by a user. (Supplementary note 8) 　　The method according to supplementary note 6, comprising: 　　Processing, by the processing unit, the updated segment specific contributions to update the forecast value in the event of change in the trend and pattern information of the at least one variable value of one or more different segments. (Supplementary note 9) 　　A system for predicting a value based on processing of one or more identified values, comprising: 　　a trained database of one or more trained identified values, controlled by a processor by applying machine learning techniques on the one or more identified values; 　　the trained database is configured to provide at least one trained identified value along with at least one variable value of one or more different segments to a machine learning model; 　　a machine learning engine, processing the trained identified value with at least one variable value for generating an estimated value for each variable segment; 　　a processing unit, processing the estimated values of each variable segment along with the trained identified values of that variable segment to create plurality of variable segment specific contributions; 　　the processing unit, generating a forecast value of at least one variable for a target period by applying specific machine learning techniques on at least one segment specific contributions. (Supplementary note 10) 　　The system according to supplementary note 9, wherein the one or more identified values are the numerical historical data values of a cashflow component and are provided by a user. (Supplementary note 11) 　　The system according to supplementary note 9, wherein the at least one variable of one or more segments is a numerical data value of a present period and is comprises of different metrics including a delay metric and cashflow components. (Supplementary note 12) 　　The system according to supplementary note 11, wherein the machine learning engine is being trained by processing the at least one variable value of one or more different segments with the trained identified values based on a trend and / or pattern information of the at least one variable value of one or more different segments, the trend and pattern information being dynamically updated by the user. (Supplementary note 13) 　　The system according to supplementary note 12, comprises: 　　an analytical module of the processing unit for analysing the effect of trend and pattern information of the at least one variable value of one or more different segments on the historical data values of a cashflow component. (Supplementary note 14) 　　The system according to supplementary note 13, wherein the plurality of variable segment specific contributions is being updated for each trend and pattern information based on change of the different metrics. (Supplementary note 15) 　　The system according to supplementary note 11, wherein the target period is set by a user. (Supplementary note 16) 　　The system according to supplementary note 14, wherein: 　　the processing unit processing the updated segment specific contributions to update the forecast value in the event of change in the trend and pattern information of the at least one variable value of one or more different segments.

[0047] This application is based upon and claims the benefit of priority from India patent application No. 202441006561, filed on January 31, 2024, the disclosure of which is incorporated herein in its entirety by reference.

[0048] 200　　SYSTEM 202　　TRAINED DATABASE 204　　PROCESSOR 206　　ML MODEL 208　　MACHINE LEARNING ENGINE 210　　PROCESSING UNIT 212　　ANALYTICAL MODULE

Claims

1. A method of predicting a value based on processing of one or more identified values, comprising: 　　applying, by a processor, machine learning techniques on the one or more identified values for generating a trained database of one or more trained identified values; 　　providing, by the trained database, at least one trained identified value along with at least one variable value of one or more different segments to a machine learning model; 　　processing the trained identified value with at least one variable value by a machine learning engine for generating an estimated value for each variable segment; 　　processing, by a processing unit, the estimated values of each variable segment along with the trained identified values of that variable segment to create plurality of variable segment specific contributions; and 　　generating, by the processing unit, a forecast value of at least one variable for a target period by applying specific machine learning techniques on at least one segment specific contributions.

2. The method according to claim 1, wherein the one or more identified values are the numerical historical data values of a cashflow component and are provided by a user.

3. The method according to claim 1, wherein the at least one variable of one or more segments is a numerical data value of a present period and is comprises of different metrics including a delay metric and cashflow components.

4. The method according to claim 3, wherein the machine learning engine is being trained by processing the at least one variable value of one or more different segments with the trained identified values based on a trend and / or pattern information of the at least one variable value of one or more different segments, the trend and pattern information being dynamically updated by the user.

5. The method according to claim 4, comprises: 　　Analysing, by an analytical module of the processing unit, the effect of trend and pattern information of the at least one variable value of one or more different segments on the historical data values of a cashflow component.

6. The method according to claim 5, wherein the plurality of variable segment specific contributions is updated for each trend and pattern information based on change of the different metrics.

7. The method according to claim 1, wherein the target period is set by a user.

8. The method according to claim 6, comprising: 　　Processing, by the processing unit, the updated segment specific contributions to update the forecast value in the event of change in the trend and pattern information of the at least one variable value of one or more different segments.

9. A system for predicting a value based on processing of one or more identified values, comprising: 　　a trained database of one or more trained identified values, controlled by a processor by applying machine learning techniques on the one or more identified values; 　　the trained database is configured to provide at least one trained identified value along with at least one variable value of one or more different segments to a machine learning model; 　　a machine learning engine, processing the trained identified value with at least one variable value for generating an estimated value for each variable segment; and 　　a processing unit, processing the estimated values of each variable segment along with the trained identified values of that variable segment to create plurality of variable segment specific contributions, wherein 　　the processing unit, generating a forecast value of at least one variable for a target period by applying specific machine learning techniques on at least one segment specific contributions.

10. The system according to claim 9, wherein the one or more identified values are the numerical historical data values of a cashflow component and are provided by a user.

11. The system according to claim 9, wherein the at least one variable of one or more segments is a numerical data value of a present period and is comprises of different metrics including a delay metric and cashflow components.

12. The system according to claim 11, wherein the machine learning engine is being trained by processing the at least one variable value of one or more different segments with the trained identified values based on a trend and / or pattern information of the at least one variable value of one or more different segments, the trend and pattern information being dynamically updated by the user.

13. The system according to claim 12, comprises: 　　an analytical module of the processing unit for analysing the effect of trend and pattern information of the at least one variable value of one or more different segments on the historical data values of a cashflow component.

14. The system according to claim 13, wherein the plurality of variable segment specific contributions is being updated for each trend and pattern information based on change of the different metrics.

15. The system according to claim 11, wherein the target period is set by a user.

16. The system according to claim 14, wherein: 　　the processing unit processing the updated segment specific contributions to update the forecast value in the event of change in the trend and pattern information of the at least one variable value of one or more different segments.

Images