53 results about "Federated search" patented technology

The invention provides a path planning method and device, electronic equipment and a storage medium, relates to the technical field of computers, in particular to the fields of artificial intelligence, intelligent transportation, automatic driving, deep learning technologies and the like, and can be used for application scenes such as path planning and intelligent driving. According to the specific implementation scheme, a transverse and longitudinal combined search result track is generated according to the vehicle state and obstacle information; generating a decision list according to the transverse and longitudinal combined search result trajectory and the obstacle information; determining a channel corridor constraint and a reference path according to the vehicle state, the obstacle information, the decision list and the environment information; and generating a planned path according to the vehicle state, the passage corridor constraint and the reference path. According to the scheme, the precision and reliability of path planning can be improved.

The invention discloses an intelligent diagnosis and risk assessment method and system for cerebral apoplexy related to atrial fibrillation, and relates to the technical field of atrial fibrillation detection.The method comprises the steps that multi-dimensional data of atrial fibrillation patients in a clinical information system are integrated, a structured database is generated, and a standardized data set is output; a standardized data set is adopted to train a first machine learning model, and model performance is optimized through parameter joint search and a training set-verification set convergence dynamic monitoring mechanism; performing cross validation on a feature weight sorting result in the optimized diagnosis model and a clinical index risk association degree calculated by a second machine learning model to generate an interaction map; and based on clinical event data containing timestamps in the structured database, adopting a third machine learning model to extract time sequence features, and combining with a survival probability analysis model to generate a risk assessment report. According to the invention, through an intelligent model adjusting and optimizing mechanism, multi-dimensional medical data are effectively integrated, and the recognition precision of the atrial fibrillation related cerebral apoplexy is greatly improved.

The embodiment of the invention relates to the technical field of unmanned vehicles, and discloses a multi-unmanned-vehicle cooperative path planning and scheduling method, device and equipment and a medium. According to the embodiment of the invention, on one hand, the joint search and global path reservation mechanism is introduced, so that the overlapping of the paths of multiple unmanned vehicles in time and space can be relieved fundamentally, the active collision prevention and congestion prevention of the global unmanned vehicles can be realized, and the system safety is improved. And on the other hand, the scheme of the embodiment of the invention can be moved up to the cloud, and the cloud control platform performs centralized decision making, so that the advantages of cloud computing can be inherited, the hardware cost of the single vehicle can be reduced, and the maximum total throughput and the shortest total task completion time of the system under multi-vehicle cooperation can be ensured (different from the target of realizing the shortest path of the single vehicle generally focused in the prior art); the global optimization value of cloud control is brought into play, and intensive and efficient utilization of computing power and road resources is realized.

The invention provides a joint retrieval method and system based on knowledge enhancement, a medium and a terminal, and ensures that a retrieval result is not limited to text segments with similar literal surfaces through a structured knowledge base, a domain label knowledge base and a multi-source heterogeneous knowledge graph. And instead, an organic knowledge whole integrating direct clauses, related law articles, judgment key items and internal rules is formed, so that the generated reply has a rigorous legal basis and a complete business context, and claim settlement disputes and legal risks caused by lack of key knowledge association are reduced. And the compliance control is changed from lagging manual inspection to real-time beforehand system built-in, so that the compliance vacuum period caused by policy update lagging is reduced, and the supervision punishment and reputation risk of an enterprise is reduced. The historical versions are reserved, and the version replacement chain is established, so that explanation based on historical contexts is provided for insurance policy life cycle query crossing different time points, and the credibility and accuracy of customer services are enhanced.

A method includes generating, by a large language model (LLM), a multitude of search queries from a user input term obtained from a user application. The method further includes, executing, by the LLM, the multitude of search queries to obtain a multitude of verified search results regarding a multitude of candidates. A multitude of web presence schemas corresponding to the multitude of candidates from the multitude of verified search results is generated. The method further includes web-scraping a multitude of websites of a subset of candidates selected from the multitude of candidates to obtain a set of corresponding web-scraping payloads. The method further includes merging the corresponding web-scraping payloads with corresponding web presence schemas of the subset of candidates to obtain a multitude of updated web presence schemas. The method further includes presenting the multitude of updated web presence schemas of the subset of candidates in the user application.

The invention provides a federal retrieval enhancement generation method oriented to multiple data sources. The method comprises the following steps: initializing the reliability of federal data sources; dividing a user query field to obtain the reliability of the federal data source to the user query field; selecting a target data source from the federated data source; performing local retrieval on the target data source to obtain candidate answers; estimating the confidence of the candidate answers; according to the confidence coefficient, filtering the candidate answers, and generating a user query answer; and according to the confidence, updating the reliability of the federal data source, and performing confidence estimation on the candidate answers, so that a confidence estimation result has rich semantic information and can be well identified and utilized by a large language model, thereby being beneficial to solving the conflict problem between exogenous knowledge and model endogenous knowledge, and improving the reliability of the model. And through federal data source quality detection of feedback iteration, reliability evaluation of the data source is continuously converged to the domain knowledge accuracy, efficient and accurate detection of the data source quality is realized, and query of invalid data sources and low-quality data sources is effectively reduced.

The invention relates to a full-process-oriented data asset model management method and device, equipment and a medium, and belongs to the technical field of data processing. The method comprises the following steps: sequentially performing metadata extraction, data asset directory construction and data asset access based on metadata association and directory binding on multiple data sources of a target business system, and constructing and generating a data asset full-flow chart to display a full-flow path of data asset access; predicting and supplementing missing metadata based on the double algorithms of data consanguinity logical reasoning and knowledge graph rule judgment; performing correctable multi-stage multi-round recursive quality evaluation on the data asset full-flow chart after missing metadata prediction and completion; and synchronously executing a three-layer parallel joint retrieval mechanism based on a quality evaluation result to complete joint search of multiple data assets and high-quality retrieval result display. By adopting the method, efficient and accurate management of the data asset model of the target business system can be realized.

The invention discloses a micro-seismic signal classification method based on genetic algorithm optimization. The method comprises the following steps: data acquisition: acquiring waveform data containing multiple types of micro-seismic event forms; preprocessing: carrying out mean value removal and band-pass filtering preprocessing on the obtained waveform data, and removing baseline drift and high-frequency noise interference; envelope calculation: respectively calculating a time domain envelope and a frequency spectrum envelope for the preprocessed microseismic signal; feature extraction: extracting a plurality of statistics and time frequency parameters from the time domain envelope and the frequency spectrum envelope to form a feature complete set; feature selection and classifier optimization: utilizing a genetic algorithm to perform joint search optimization on structure parameters and feature subsets of the multi-layer perceptron; and training and verification: training the multilayer perceptron by using the optimized feature subset and the network structure, and evaluating the classification accuracy on a verification set. The method can realize high-precision classification, has robustness to noise, and can be widely applied to the fields of microseism monitoring, blasting event monitoring, other vibration source identification and the like.

The invention provides a double-server combined keyword dynamic searchable encryption method with forward and backward security, which is applied to a system comprising a data owner, a data user, a first server and a second server. Keywords and associated file identifiers are stored in each leaf node, information of non-leaf nodes is mapped to a non-distinguishable long filter, and through a random state updating mechanism, a keyword double-state array and a time factor trap door design, it is guaranteed that keyword joint search is achieved, and meanwhile, the information of the non-leaf nodes can be quickly and efficiently searched. And the forward safety and the backward safety are also ensured. According to the dual-server architecture, security index storage and trap door verification are separated, so that the problem of KPRP leakage is effectively relieved.

Apparatus and methods of joint search space set for enhanced PDCCH transmission with multiple beams from multiple TRPs are disclosed. The apparatus includes: a processor that generates a plurality of versions of Downlink Control Information (DCI) for transmission from a plurality of transmitting-receiving identities with a plurality of Control Resource Sets (CORESETs), each CORESET being transmitted from one of the transmitting-receiving identities; and configures an Information Element (IE) indicating a joint search space set for the DCI, wherein the joint search space set comprises a plurality of component search space sets; and a transmitter that transmits the DCI and the IE using the plurality of transmitting-receiving identities; wherein each one of the component search space sets is configured for monitoring and detecting one DCI version.

The invention provides a link prediction method, system and device for a graph network and a storage medium, and belongs to the crossing field of a graph neural network and machine learning, and the method comprises the steps: firstly, constructing a three-dimensional joint search space composed of a sub-graph extraction strategy, a motif structure type and a graph neural network architecture, and through a differentiable architecture search technology, carrying out the three-dimensional joint search of the sub-graph extraction strategy, the motif structure type and the graph neural network architecture; automatically learning an optimal analysis combination for a specific network structure; the optimal analysis combination is used for prediction, the system fuses local adjacency information based on the self-adaptive subgraph and a high-order structure mode based on the selected motif, the optimized graph neural network is used for coding and aggregation, and finally the high-precision link existence probability is output. Experimental verification shows that the method achieves remarkable and stable performance improvement on isomorphic and heterogeneous data sets, the limitation of a traditional method under different topological structures is effectively overcome, and a universal, robust and efficient solution is provided for a link prediction task in a complex network.

The application provides a kind of generation of adversarial network architecture search method based on multi-population coevolution, comprising: step 1, define the search space for searching generator and discriminator;Step 2, randomly generate the gene population of generator and discriminator;Step 3, activate all candidate operations, and initialize training;Step 4, search the architecture parameters and network weights of generative adversarial network;Step 5, optimize and update the gene population of generator and discriminator;Step 6, alternately optimize the architecture of generator and discriminator until the network performance can not be improved any more;Step 7, the searched generative adversarial network is used for the generation of unseen samples in zero-shot learning.The joint search of the method can fully utilize the interdependent and dynamic game relationship between generator and discriminator in GAN framework, which is conducive to searching better GAN architecture, and has higher efficiency and better search performance compared with traditional optimization methods.

The invention discloses a same-frequency multi-target signal detection method for Doppler-code phase two-dimensional joint search, and relates to the field of spaceflight measurement and control. Through Doppler frequency shift generated by satellite motion and code phase deviation caused by signal propagation time delay difference, two-dimensional joint correlation calculation and incoherent accumulation are carried out at a receiving end to generate a global energy matrix, an adaptive threshold is calculated, strong signal side lobe interference is suppressed by defining a protection area, the weak signal detection probability is increased to 90% or above, and the detection accuracy is improved. Tracking channels are distributed according to an energy descending order, and the purpose of simultaneously and efficiently detecting multiple targets is achieved. The method is particularly suitable for same-frequency multi-target signal detection based on Doppler-code phase two-dimensional joint search in a low-orbit satellite constellation or a dense satellite network, and solves the problems of low spectrum utilization rate, complex pseudo code management and the like of the traditional frequency division / code division technology.

The application discloses a noise elimination and transformation model robust estimation method in multi-modal image registration, comprising: obtaining initial matching point pairs, establishing a homography transformation model to be estimated and residual error measurement; performing vector field consistency calculation on the initial matching point pairs, and constructing an edge expansion consensus set; performing initial fitting on the homography transformation model on the edge expansion consensus set to obtain an initial homography transformation model; estimating noise scale and initializing residual error threshold based on residual error statistics induced by the initial homography transformation model; regarding the residual error threshold and the homography transformation model parameters as same-order decision quantities, constructing a joint optimization framework of threshold-model cooperation; adaptively constructing a joint search domain according to regularity statistics of residual error distribution; constructing a threshold-coupled cost function under a mixed statistical modeling framework of inliers / outliers; performing collaborative iterative updating on the homography transformation model and the residual error threshold by adopting a group optimization strategy, and outputting the homography transformation model and a refined inlier set.

The invention discloses a multi-task joint search method and system, a medium and an intelligent cockpit, and the method comprises the steps: recognizing a multi-target cooperation task which meets the requirements of a user from the inquiry information of the user, and disassembling the multi-target cooperation task into a plurality of single-target tasks with a dependency relationship; searching the plurality of single-target tasks to obtain candidate object sets corresponding to the plurality of single-target tasks; aiming at the current single-target task, performing arbitration scoring on each candidate object in the candidate object set corresponding to the current single-target task by utilizing a comprehensive arbitration model, and screening a target candidate object corresponding to the current single-target task by referring to an arbitration score; wherein the comprehensive arbitration model is at least associated with an environment factor to which the user inquiry information belongs to carry out arbitration scoring on the matching degree of the candidate object; and according to the dependency relationship of the plurality of single target tasks, combining the target candidate objects corresponding to the plurality of single target tasks to obtain a reply aiming at the user inquiry information.

The present disclosure relates to systems and methods for generating contextually, grammatically, and conversationally correct answers to input questions. Embodiments provide for linguistic and syntactic structure analysis of a submitted question in order to determine whether the submitted question may be answered by at least one headnote. The question is then further analyzed to determine more details about the intent and context of the question. A federated search process, based on the linguistic and syntactic structure analysis, and the additional analysis of the question is used to identify candidate question-answer pairs from a corpus of previously created headnotes. Machine learning models are used to analyze the candidate question-answer pairs, additional rules are applied to rank the candidate answers, and dynamic thresholds are applied to identify the best potential answers to provide to a user as a response to the submitted question.

The present disclosure provides a trajectory planning method, device, equipment, storage medium and automatic driving vehicle, which relates to the fields of automatic driving, intelligent traffic and the like, and can be used in application scenarios such as vehicle path planning. The specific scheme is: determining obstacle information according to environment information, and constructing a path search environment and a joint search environment respectively; in the path search environment, performing path search based on a preset steering operation node, determining direction guide information and correcting an initial reference line and an initial search corridor to obtain a reference line and a search corridor; performing space-time joint search in the joint search environment according to the reference line and the search corridor to obtain an initial space-time trajectory, and generating interaction decision information; determining boundary constraints and optimizing the initial space-time trajectory according to the interaction decision information to generate a decision planning trajectory. The present scheme can uniformly process path planning and speed planning, and improve the consistency, executability and stability of trajectory output in a complex road environment.

A data intake and query system can process a query to identify subquery tokens corresponding to subqueries to be executed by external data systems. The data intake and query system can process the subquery tokens to generate modified subqueries to be executed by the external data systems. The modified subqueries can cause the external data system to return metadata associated with the events processed by the external data systems during executing of the modified subqueries.

A parallel and distributed query engine for federated searching is disclosed herein. As contemplated by the present disclosure, the system may provide a single application programming interface that allows a user to access and analyze multiple enterprise data storage locations remotely and simultaneously while presenting and reporting information from the multiple sources in a single, uniform display. Such a solution may allow a user to analyze and cross-reference data stored in multiple locations by using multiple queries in real time without requiring the actual data files to be displaced or combined. The system may further implement interactive artificial intelligence assistant, natural language processing, and workflow-based operations for improved user access and functionality.

The present disclosure relates to systems and methods for generating contextually, grammatically, and conversationally correct answers to input questions. Embodiments provide for linguistic and syntactic structure analysis of a submitted question in order to determine whether the submitted question may be answered by at least one headnote. The question is then further analyzed to determine more details about the intent and context of the question. A federated search process, based on the linguistic and syntactic structure analysis, and the additional analysis of the question is used to identify candidate question-answer pairs from a corpus of previously created headnotes. Machine learning models are used to analyze the candidate question-answer pairs, additional rules are applied to rank the candidate answers, and dynamic thresholds are applied to identify the best potential answers to provide to a user as a response to the submitted question.

The invention provides a joint search method and device for a neural network and hardware, is applied to the technical field of computers, and is used for solving the problem that defects exist in the aspects of precision evaluation and hardware performance search in the prior art. The method comprises the following steps: sampling a neural network search space corresponding to a current round of iteration to obtain G to-be-evaluated neural networks corresponding to the current round of iteration, and determining a precision score value of the gth to-be-evaluated neural network; performing hardware performance simulation on the hardware configuration parameters of the to-be-evaluated neural networks by using the heterogeneous multi-core architecture, and determining a hardware performance value of the gth to-be-evaluated neural network; according to the ratio of the precision score value to the hardware performance value of the gth to-be-evaluated neural network, determining the adaptability value of the gth to-be-evaluated neural network, and determining a candidate neural network corresponding to the current round of iteration in the G to-be-evaluated neural networks; and determining a target neural network suitable for executing the target processing task from the multiple rounds of respective corresponding candidate neural networks.

The invention discloses a model construction method based on training-free neural architecture search and an automatic deep learning platform, and relates to the technical field of artificial intelligence. The model construction method based on training-free neural architecture search comprises the following steps: acquiring uploaded data of a user, and labeling the uploaded data; the annotated uploaded data are input into a search module based on an NAS-HPO combined search space; a search module based on the NAS-HPO joint search space simultaneously optimizes the neural architecture parameters and the hyper-parameters to generate one or more candidate neural networks; evaluating one or more candidate neural networks in a training-free manner, and determining a Top-1 network in the candidate neural networks; the Top-1 network is finely tuned; and packaging the Top-1 network after fine tuning into a deployable software package. By adopting the technology provided by the invention, the computing power resource can be effectively saved, the construction speed is accelerated, and the full-process automation is realized to reduce the requirement on the professional degree of a user.

The invention provides a signal detection method, apparatus and device, and a storage medium, which can effectively avoid the problem of self-information interference without introducing iterative feedback and feeding back the estimation result of the symbol for self-correction by constructing a state based on an original received symbol and a known channel structure, and improve the detection efficiency of the signal. An interference structure with any complexity can be accurately described, high-precision symbol detection is realized, and the method is suitable for any time delay-Doppler extension scene. In the process of advancing from the state of one received symbol to the state of another received symbol, the differentiated undetermined sending symbols playing a key information updating role between adjacent states are concerned, all sending symbols do not need to be reprocessed, joint search for full combinations is avoided, and the transmission efficiency is improved. The calculation complexity is reduced from the index level to the polynomial level related to the difference scale between the states, and the processing capacity is improved on the premise that the detection performance is not reduced.

The application discloses a kind of based on defining organization attack preference guard graph decision method and system, its method includes steps: from the log server of honeypot collection attack information, through the joint search of threat intelligence knowledge graph and external threat intelligence library, ascription defines the behavior mode of attack organization, and the attack preference of attack organization is sorted;From the perspective of attacker, through the topological information in target network, asset information and configuration information generation attack graph, and the mapping of defining organization attack preference to attack graph node is carried out, generates guard graph, and then generates the guard matrix for quantifying calculation attack graph attack mode preference influence;Based on the analysis of the guard matrix of attack graph to defining organization attack preference, with emphasis, generate honeypot service, provide decision support for the deployment of defense measures.The application can more flexible quickly generate the type of honeypot service, greatly saves manpower and time cost for the analysis and deployment of deception defense.

A data intake and query system can process a query to identify subquery tokens corresponding to subqueries to be executed by external data systems. The data intake and query system can process the subquery tokens to generate modified subqueries to be executed by the external data systems. The modified subqueries can cause the external data system to return metadata associated with the events processed by the external data systems during executing of the modified subqueries.

The application discloses a Doppler-code phase two-dimensional joint search same-frequency multi-target signal detection method and relates to the field of spaceflight measurement and control. The application generates a global energy matrix through two-dimensional joint correlation calculation and incoherent accumulation at a receiving end according to the Doppler frequency shift caused by satellite movement and the code phase offset caused by signal propagation time delay difference, calculates an adaptive threshold, suppresses strong signal sidelobe interference by defining a protection area, and improves weak signal detection probability to more than 90%. The application allocates tracking channels in descending order of energy, and achieves the purpose of efficient detection of multiple targets at the same time. The application is especially suitable for Doppler-code phase two-dimensional joint search same-frequency multi-target signal detection in a low-orbit satellite constellation or a dense satellite network, and solves the problems of low spectrum utilization and complex pseudo code management of traditional frequency division / code division technology.

The application provides a gait emotion joint recognition method based on heterogeneous space neural architecture search, comprising the following steps: extracting three-dimensional human posture joints in different emotion video data; performing space conversion, and converting into a rectangular coordinate system and a spherical coordinate system respectively; performing position information space mapping and angle information space mapping on the joints in different coordinate systems based on a similarity minimization strategy; constructing a neural architecture search space and a joint search strategy in different spaces, searching for optimal backbone networks respectively, and training the optimal backbone networks respectively; extracting human time sequence gait emotion features in different spaces by using the trained optimal backbone networks, inputting the features into a Softmax classifier, and obtaining scores of different emotions; and performing joint scoring to complete the gait emotion joint recognition based on the heterogeneous space neural architecture search.

The present invention belongs to the field of power technology, and particularly relates to an optimized scheduling method for a combined cooling, heating and power integrated energy utilization system, comprising the following steps: S1, obtaining the predicted power generation of photovoltaic and wind power, the predicted cooling, heating and power loads of users, and electricity price information during the optimization period; S2, taking the time-of-use electricity price, wind power, predicted wind power generation and load during each optimization period as the inputs of the system, using the gas turbine, conventional generator set, standby generator set and battery output as decision variables, and obtaining the optimal output of each generator set at each time period through the PSO-CS joint search algorithm. The present invention is used for the optimized scheduling of a combined cooling, heating and power integrated utilization system. Through the heuristic algorithm, the problems of multiple types of system equipment, complex coupling and difficult scheduling can be effectively solved, thereby improving the economic operation of the system. By combining the particle swarm optimization algorithm with the cuckoo algorithm, the problem that the particle swarm optimization algorithm is prone to falling into a local optimal solution in the later stage is solved.

The invention discloses a noise elimination and transformation model robust estimation method in multi-modal image registration, and the method comprises the steps: obtaining an initial matching point pair, and building a to-be-estimated homography transformation model and a residual measurement; vector field consistency calculation is carried out on the initial matching point pairs, and an edge expansion consensus set is constructed; performing initial fitting on the homography transformation model on the edge expansion consensus set to obtain an initial homography transformation model; estimating a noise scale based on residual statistics induced by the initial homography transformation model and initializing a residual threshold; regarding a residual threshold and a homography transformation model parameter as same-order decision quantities, and constructing a threshold-model collaborative joint optimization framework; a joint search domain is constructed in a self-adaptive mode according to regularity statistics of residual error distribution; constructing a threshold coupling cost function under an inner point / outer point mixed statistical modeling framework; and carrying out collaborative iteration updating on the homography transformation model and the residual threshold by adopting a group optimization strategy, and outputting the homography transformation model and a refining inner point set.

This application relates to automatic operation detection on protected fields that support federated search. Systems and methods are provided for automatic operation detection on protected fields. A data model configuration can be used to specify which properties of a data model used by a cloud-based application are protected by a data security provider that monitors communications between the application and client devices. A determination can be made automatically as to which operations of the cloud-based application are supported for protected fields. The cloud-based application can be configured to enable / disable certain features, such as validators, autocomplete, search operators, etc., depending on whether a property is a protected field.