129 results about "Algorithm Selection" patented technology

The invention discloses a QoS guarantee method and system for a communication network, and relates to the technical field of communication networks, and the method comprises the steps: collecting and preprocessing network state data, and forming a standardized data set; dynamically classifying service types based on an improved random forest algorithm, and predicting a future QoS demand trend of each priority service in combination with an LSTM neural network; establishing a mapping model of QoS demands and resource parameters, converting predicted demands into allocable resource indexes, monitoring the resource utilization rate in real time, and setting an elastic reservation mechanism and conflict early warning; when early warning is triggered, selecting an optimal transmission link by adopting a multi-path collaborative algorithm, and implementing differentiated resource allocation according to service priorities; and a closed-loop feedback mechanism is triggered to dynamically adjust resource allocation by monitoring the deviation between the actual QoS and a predicted value in real time. The method has the advantages that through multi-dimensional perception, LSTM prediction, dynamic resource management and multi-path scheduling, QoS requirements of services with different priorities are accurately matched, and dynamic changes of the network are efficiently coped with.

Disclosed in the present invention are an edge-deployed semi-supervised anomaly detection method and system for a railway track foreign object. The method comprises the following steps: an edge device encoding and decoding a video stream captured by a camera to obtain an image frame sequence, and performing frame extraction; and using a semantic segmentation model to perform image segmentation on a certain image frame obtained by means of frame extraction, to obtain a railway track region segmentation image. The use of a single image as input may generate an expert model result having a high weight value; however, the determination based on a single image is not stable, multiple consecutive images of the task scene need to be inputted, the frequency of each expert model obtaining the highest weight is computed, and the expert model corresponding to the highest frequency is the final solution. The present invention supports scene-adaptive foreign object detection algorithm automatic selection, and a user can perform selection on the basis of prior knowledge, or selection may be performed by a scene-adaptive automatic algorithm selection method; the user only needs to provide a batch of image data of the current scene, and the optimal algorithm selection can be evaluated.

The invention provides a transformer area line loss intelligent algorithm decision-making method based on a knowledge graph and a large model agent, semantic retrieval in the algorithm calling process is supported, dominant and implicit knowledge such as expert experience, technical standards and historical cases is coded in a unified mode, a mapping relation is established between algorithm nodes and process nodes in the knowledge graph, and the algorithm is optimized. An integrated reasoning path is formed, so that the intelligent agent can automatically combine and call an algorithm according to a task target, a large model is embedded into a transformer area line loss analysis process, and a hierarchical structure of'task analysis intelligent agent-algorithm decision intelligent agent-result analysis intelligent agent 'is formed; the task analysis agent is responsible for converting a problem input by a user into a structured task; the algorithm decision agent is responsible for algorithm selection and adaptive parameter adjustment; the result analysis agent is responsible for report generation and decision recommendation based on the knowledge base; according to the method, organized association can be carried out on scattered algorithm tools, expert experience and business rules, and a unified transformer area line loss agent tool integration framework is constructed.

The invention relates to the technical field of IPv6 self-configuration, in particular to an IPv6 network mode self-adaptive configuration method and system based on an RA message. In the method, a finite-state machine establishes a state transition path according to the combination relation of M and O flag bits in an RA message cache set and prefix field content, converts a network mode recognition process into a quantifiable state judgment sequence, and sends the state judgment sequence to the RA message cache set; configuration ambiguity caused by flag bit fluctuation in traditional static judgment is eliminated, pattern recognition under network environment changes has continuity and self-consistency, the classification regression tree constructs a branch judgment path with the duration, the advertisement interval and the prefix consistency of an RA source as characteristic variables, an optimal node is selected through a hierarchical splitting algorithm, and the judgment accuracy is improved. The dynamic mapping of stateful and stateless configuration paths is realized, the misjudgment rate of configuration branches is reduced, and the address allocation delay is optimized, so that the IPv6 self-configuration has a structured decision-making capability and a parameter mapping self-optimization characteristic, and the connection reliability, the address generation speed and the configuration accuracy are improved.

The invention provides a video coding and compression transmission method for intelligent equipment, which belongs to the technical field of video coding and compression transmission, realizes advanced channel quality estimation by establishing a channel prediction model enhanced by a spatial transformation network, configures layered coding parameters according to a prediction result and optimizes motion search in combination with inertial measurement unit data. An optimal coding mode is selected by an application rate distortion optimization algorithm, enhancement layer sending is adjusted according to real-time feedback by adopting a dynamic transmission scheduling strategy, and differential forward error correction protection and a time-space domain error concealment mechanism are matched; the technical problem that video transmission is unstable due to violent fluctuation of channel quality when intelligent equipment moves and works in complex electromagnetic environments such as a transformer substation is solved.

The invention discloses an automobile industry code quality automatic compliance method based on a large language model. The method comprises the following steps: calling a static code analyzer to generate a violation report; the method comprises the following steps: intelligently classifying violation into Tier1, Tier2 and Tier3 based on a multi-feature analysis algorithm of a rule document; for Tier1 violation, performing three times of retry iteration repair by using a large language model, automatically detecting compilation errors and performing rollback; for Tier2 violation, multiple technical schemes are extracted by a large language model for a user to select; a multi-LLM antagonistic cooperation mechanism is adopted for Tier3 violation, automatic restoration of complex violation is achieved through a Generator generation restoration scheme, a Disscriminator review challenge scheme and antagonistic iteration of Arbiter arbitration conflicts, and an optimal scheme is selected through a weighted voting algorithm when conflicts occur. The method obviously reduces the manual intervention rate, improves the repairing quality, and is low in cost and high in universality.

The invention discloses a two-stage algorithm selection and hyper-parameter joint optimization method, which comprises the following steps of: in the first stage, processing a training set and a test set through row sampling operation and column dimension reduction operation to form a reduced data set; randomly sampling a certain number of configurations in the hyper-parameter space of each candidate algorithm, evaluating the performance of each candidate algorithm by using the reduced data set, and extracting an optimal performance score; in the second stage, a previous algorithm is screened according to the optimal performance score to form a candidate set, and a pruned hyper-parameter search space is formed so as to reduce the calculation complexity of processor hyper-parameter search; and performing hyper-parameter optimization on the pruned hyper-parameter search space by using the original data set, and outputting an optimal algorithm adaptive to the target technical task and hyper-parameter configuration thereof. Algorithm screening and hyper-parameter tuning adaptive to a specific scene are realized through a two-stage optimization strategy, and meanwhile, the method is suitable for a traditional table type dichotomy task and aims at improving the deployment efficiency and performance of a machine learning model.

The invention belongs to the technical field of computer vision and three-dimensional reconstruction, and particularly relates to a building intelligent three-dimensional reconstruction method and system based on point cloud, and the method comprises the steps: obtaining the point cloud data of a building, and carrying out the preprocessing of the point cloud data; constructing an intelligent three-dimensional reconstruction model based on an intelligent algorithm selection mechanism, an adaptive surface reconstruction technology and an anti-mold-penetration processing algorithm; and three-dimensional reconstruction of the building is realized based on the preprocessed point cloud data and the intelligent three-dimensional reconstruction model. Through an intelligent algorithm selection mechanism, a self-adaptive surface reconstruction technology and an anti-mold-penetration processing algorithm, the technical problems of blind algorithm selection, unstable reconstruction quality, geometric mold penetration and the like in the prior art are solved, and high-quality and high-efficiency building three-dimensional reconstruction of point cloud data of various formats is realized.

The invention provides a zero-trust-based dynamic data secure transmission and processing channel construction device and method, and the method comprises the steps: S1, collecting to-be-transmitted data, and classifying the data to obtain a data classification result; s2, based on the data classification result, the user identity information and the environmental risk information, a dynamic security policy is generated, and the dynamic security policy comprises an encryption algorithm selection rule, a trust evaluation threshold and a data operation authority policy; according to the invention, the fine-grained security policy is generated and continuously adjusted by dynamically sensing the data sensitivity and the real-time trust state, so that the full-link self-adaptive security protection from transmission to processing is realized; according to the method, the security multi-party calculation is seamlessly integrated in a high-sensitivity scene, so that the end-to-end security and privacy protection level of data in a cross-domain and multi-participant environment are remarkably improved while the data mobility is guaranteed, and the problem that a traditional static security mechanism is difficult to adapt to dynamic risks is effectively solved.

The invention relates to the technical field of computer vision and deep learning, in particular to a signature and seal verification method and system based on target recognition. The method comprises the following steps: collecting signature and seal image data in an image file, and generating a structured data set containing category labels and position information; processing the structured data set; a convolutional neural network is adopted to construct a feature extraction network, local features of a target are captured through a convolutional layer, downsampling dimensionality reduction is realized through a pooling layer, and high-dimensional features are mapped through a full connection layer; a YOLO algorithm and an SSD algorithm are introduced as target identification core algorithms; setting unified training parameters and evaluation indexes, and training and testing the optimized YOLO algorithm and SSD algorithm; the system comprises a data set construction module, a data enhancement module, a feature extraction module, an algorithm selection and optimization module and an experiment verification module. The purpose of improving the accuracy and efficiency of signature and seal verification is achieved.

Systems and processes are disclosed for a multi-level quantum-based vertically classified entropy exploratory analytics tool designed to improve speed, accuracy, and scalability in anomaly detection and data analysis. The tool employs a dynamic algorithm selector for adaptive algorithm choice, a quantum encoder for precise data encoding, and a multi-level splitter and aggregator for efficient data segmentation and result integration. It includes a classification executor for accurate decision-making, an exploratory data analyzer for uncovering hidden patterns, and a multi-dimensional data processor for handling complex data sets. A qubit selector optimizes quantum resource allocation. The tool combines classical and quantum computing methods, enhancing robustness and versatility. This system significantly reduces false positive rates and improves processing efficiency, addressing the limitations of classical methods in handling large-scale, multi-dimensional data sets. The invention is particularly valuable for applications requiring rapid and precise data analysis, such as finance, cybersecurity, and scientific research.

The invention discloses a mine emergency data relay transmission method based on LoRa and 4G fusion, and belongs to the technical field of mine communication. The method comprises the steps that a communication node monitors sensor data in real time, and when key emergency data is recognized, 4G signal strength is automatically detected; when the 4G signal strength is lower than a set threshold value, starting a LoRa multi-hop relay transmission mode, and selecting an optimal relay node through a dynamic relay node discovery and comprehensive preferential algorithm; the key data is transmitted to an exit node with 4G connection capability through a one-hop or multi-hop LoRa link, and is finally uploaded to a command center through a 4G network. According to the invention, through advantage complementation of LoRa and 4G, a hybrid communication network adaptive to a complex mine environment is constructed, and the technical problem of reliable transmission of key data in a communication blind area is effectively solved.

The invention discloses a data modeling system based on research data, which relates to the technical field of data modeling and comprises a research data acquisition module, a research data cleaning and conversion module, a research data encryption module, a model building, screening and optimizing module and a modeling algorithm library. According to the method, dynamic model construction algorithm selection can be carried out according to research data, model construction processing is carried out, the flexibility and diversity of research data modeling attachment forms are improved, the research data are finely analyzed, reasonable screening of the model construction algorithms can be carried out according to specific data characteristics, and the model construction efficiency is improved. According to the method, the intelligence and coordination of model optimization processing are powerfully supported and guaranteed, higher convenience is also provided for calling and checking of subsequent models, encryption processing is performed on the structured data set by using the AES symmetric encryption algorithm, the attention on data transmission and storage security is improved, and the method is suitable for being used in the field of data transmission and storage. And reliable guarantee can be provided for safe storage of data.

This invention discloses a short-term power load forecasting method based on KNN federated distillation learning and Seq2Seq. The method involves acquiring historical power load data for a preset time period, inputting it into a trained local model on an edge computing device, and obtaining power load data for a future preset time period. This method trains local models from multiple edge computing devices together, employing federated learning and knowledge distillation during training. The local models use an encoder to extract spatiotemporal features and a decoder to convert these features into output for prediction. The KNN algorithm selects the parameters of the most similar local models from each edge computing device and aggregates them as the parameters of the teacher model. Knowledge distillation is achieved by transmitting knowledge between edge computing devices, rather than simply using a global model to update local models, thus reducing the impact of data heterogeneity. Experimental results show that the proposed S3TKFDL has better performance and robustness compared to the baseline model.

The application relates to the technical field of data security storage, and discloses a data encryption storage method and system of a storage-evidence platform. The method acquires an original data packet of to-be-encrypted storage-evidence data, identifies data type marks and sensitive field distribution characteristics therein. According to the identification result, a multi-layer encryption domain model is constructed, the original data packet is divided into multiple data segments and is distributed to corresponding encryption domain levels. A dynamic encryption strategy containing a key generation rule, an encryption algorithm selection rule and a data confusion rule is independently generated for each encryption domain level. According to the strategy, a hierarchical encryption operation is performed on each data segment, and an encrypted data segment set with a hierarchical check mark is generated. All encrypted segments are recombined into a primary encrypted data block in hierarchical order, and an encryption path index is embedded. Access behavior characteristics of the primary data block in a storage node are collected, and a behavior-triggered secondary encryption instruction is generated based on the characteristics.

The invention discloses an intelligent control method and system for a heating station based on reinforcement learning, belongs to the field of intelligent control of a heating system, and aims to solve the problems that an existing control method and system for the heating station cannot give consideration to the comfort of a user and the energy-saving performance of the system at the same time, and system fluctuation is caused by insufficient heat supply, excessive heat supply and frequent adjustment frequently. The method provided by the invention comprises the following steps: building a data acquisition module, building a data preprocessing module, building a feature engineering module, building a data set division module, building a reinforcement learning modeling module, building a reinforcement learning algorithm selection and network structure design module, and building a model training and deploying module. The provided system corresponds to the method, the method and the system intelligently regulate and control the heating station, dynamic balance of user comfort and system energy efficiency is achieved, dynamic working conditions are adapted, regulation and control are fine and smooth, and stable operation of the system is guaranteed.

The application relates to the technical field of data security, in particular to a data desensitization method and system based on the combination of an MCP protocol and rule driving. The method comprises the following steps: loading a business rule through an MCP Server and registering a desensitization tool as a standardized tool; receiving a user query through an MCP Client and analyzing a user intention by an AI intelligent agent; dynamically calling the standardized tool to form a desensitization task chain according to the user intention and the business rule by the AI intelligent agent; and processing target data by executing the desensitization task chain, wherein the calling process comprises dynamic strategy adjustment based on context information. The system comprises an MCP service layer, a desensitization tool library module and an intelligent decision and integration layer. The application realizes dynamic weight distribution and probabilistic algorithm selection of a desensitization strategy through MCP protocol standardization integration, business rule driving and AI intelligent agent decision, effectively improves the flexibility, intelligent level and system integration efficiency of data desensitization, and simultaneously strengthens the safety and compliance guarantee capability.

This application discloses a key distribution generation method and system for multi-park integrated authentication, relating to the field of industrial internet key technology. The method includes: collecting cardholder information and sending it to a cloud center; generating a first authentication risk coefficient based on the current operating status and a second authentication risk coefficient based on historical verification records, and evaluating the authentication risk index; if the authentication risk index does not exceed a preset risk threshold, using a basic key distribution generation strategy for identity verification; if the authentication risk index exceeds the preset risk threshold, calculating the risk index overflow value, optimizing the key distribution factor composition and the selection method of the derived algorithm, and generating an adapted key distribution generation strategy for identity verification. This solves the technical problems of existing key distribution generation strategies being fixed and rigid, unable to adaptively adjust to the dynamic risk characteristics of multi-park authentication scenarios, and lacking a risk quantification assessment mechanism, leading to a lag between security strategies and risk situations.

The application discloses a kind of multi-beam outlier automatic filtering methods combined with uncertainty and density clustering algorithm, belong to multi-beam sounding data processing technical field.The application constructs grid node absorbable water depth point selection model according to CUBE filtering algorithm, and gridding is carried out to sounding data;The one-dimensional clustering of node absorbable water depth value is carried out using DBSCAN clustering algorithm, a plurality of water depth hypotheses are constructed according to the difference between water depth values, then the water depth hypothesis and uncertainty of node are updated using Kalman filter;Finally, the water depth hypothesis with higher reliability is selected from a plurality of water depth hypotheses as the real water depth value of node.The application can provide a kind of multi-beam sounding data automatic cleaning algorithm selection for domestic multi-beam sounding data processing software.

The invention discloses a computer network management method, which relates to the technical field of computer network management, and is characterized in that three groups of algorithm units in a computing processing module are matched with one another to jointly form a core architecture of the computer network management method; a plurality of influence factors such as bandwidth usage amount, data packet quantity and data packet size can be comprehensively considered, real-time network performance indexes of different users are calculated, a task with the shortest execution time is selected to be executed preferentially according to the network performance indexes and a short job priority (SJF) algorithm, network resource distribution of the multiple users is completed, and the user experience is improved. The efficiency of network resource allocation is improved, and the fairness of network resource management and allocation is ensured; and when the user carries out secondary broadband utilization rate calculation, the network performance index calculated last time is substituted into the formula to carry out iterative calculation, a new network performance index is obtained, and network resource allocation is optimized and adjusted in real time according to the new network performance index.

The information processing device acquires learning data, which includes: structural information related to the structure of the system, required performance information related to the action of the object device, setting information related to the setting of the evaluation function, calculation method selection information regarding the selection of a parameter calculation method from multiple parameter calculation methods, and evaluation information related to the evaluation results of the servo motor's driving performance on the object device. By using machine learning with multiple learning data, an estimation model is generated that estimates at least one of the optimal setting of the evaluation function and the optimal selection of the parameter calculation method based on the structural information and required performance information.

The application relates to a multi-level sentiment classification method for network public opinion. The method realizes real-time acquisition and processing of social media data through a distributed message queue, generates an initial topic set by using a Streaming K-means algorithm for online clustering, constructs a topic state space by using a topic state analysis model, selects a suitable classification level by using a Q-learning algorithm, and constructs an information entropy value dataset by using dependency path analysis and information entropy value calculation. Finally, a multi-level sentiment classification result is generated by using a level weight prediction model and a sentiment classification model, which is used for network public opinion monitoring. The method can adapt to changes in network public opinion in real time, and improves the accuracy and timeliness of sentiment classification.

A method for controlling processing of a biological material includes receiving at least one input parameter associated with the biological material: retrieving, by a data retrieval module, a first data set from a data store based on the at least one input parameter; selecting, by an algorithm selection module, a first machine learning algorithm for processing the at least one input parameter from a plurality of candidate machine learning algorithms, the first machine learning algorithm being selected based on the at least one input parameter or the first data set; training the first machine learning algorithm using at least a portion of the first data set; applying a trained first machine learning algorithm to the at least one input parameter to determine a first processing parameter for subsequent processing of the biological material; and outputting the first processing parameter for subsequent processing of the biological material.

Embodiments of the present application disclose a wearing detection method and device, a wearable device and a computer readable storage medium, and relate to the technical field of wearing detection; the method comprises: scanning sensors currently available for the wearable device, and obtaining configuration information of the sensors; inputting the configuration information of the sensors into a pre-trained algorithm selection model to obtain a target algorithm output by the algorithm selection model; the target algorithm has an accuracy rate and power consumption for wearing detection based on data collected by part or all of the sensors satisfying a preset condition; a target sensor corresponding to the target algorithm is determined, and target data collected by the target sensor is obtained; and the target data is processed based on the target algorithm to obtain a wearing detection result of the wearable device. Thus, the present application can accurately and low-power detect wearing of various wearable devices at low research and development cost.

The invention provides a data center extension planning method, system and device and a medium, and relates to the field of sustainable data centers, and the method comprises the steps: obtaining a data processing demand, dividing a region map into a plurality of data center site selection regions, and obtaining the temperature characteristics of different site selection regions; inputting the temperature characteristics into a preset electricity-water demand model based on temperature sensing to obtain power consumption demands and cooling water demands corresponding to different site selection areas; evaluating a site selection area based on the data processing demand, the power consumption demand and the cooling water demand to obtain a primary site selection area; and uncertain factors in data center planning are identified and classified, and an optimal site selection area is selected as site selection of the data center based on the uncertain factors, the parameter column and a constraint generation algorithm. According to the method, the modified thermodynamic formula is introduced, the temperature-dependent multi-system coupling relation is quantified, the demand model can be dynamically matched with the actual operation working condition, and the multi-system cooperation flexibility is remarkably improved.

Machine learning systems and methods for automated claims data and workflow processing are provided. The system includes a property claims processing platform which monitors for one or more automation trigger events in connection with an insurance claims property file. When the one or more automation trigger events is detected, the system identifies and executes one or more automated claims processing workflows in response to the detected trigger events. The workflows can be pre-defined and / or user-defined workflows that are selected in response to pre-defined logic or heuristic rules, and / or they could be selected by one or more machine learning algorithms.

The invention discloses a virtual bit configuration and VIP priority queuing method and device. The method comprises the innovative technologies of inserting virtual bits in a conventional queue according to specific intervals, calculating the optimal interval distance, dynamically adjusting virtual bit configuration, allocating fractional virtual bits and the like. According to the method, firstly, a basic rule of inserting one virtual bit after N common users are analyzed and determined through a queue model, the optimal spacing distance is calculated based on queue data, and the optimal configuration is selected by adopting weighted comprehensive scoring and a graph theory maximum cut value algorithm. The system also introduces a virtual bit dynamic adjustment mechanism, and recalculates the VIP service demand and resource configuration according to the queue change. By accurately monitoring the utilization rate of the virtual bit, automatic balance is realized between the resource insufficiency queue and the excessive queue, the queuing experience of the medical institution is remarkably improved, and the medical resource utilization efficiency and the service fairness are improved.

A novel, two-stage Distributed Fiber Optic Sensor (DFOS) placement strategy and method that ensures resilient monitoring of critical infrastructure during electrical power supply failures. This strategy and method combines a heuristic algorithm, PURE (Power Source-aware Route Exploration), with Integer Linear Programming (ILP) optimization. The PURE algorithm explores potential DFOS routes while explicitly considering the dependency of DFOS devices on the electrical power distribution network, ensuring routes electrically powered by the same electrical power feeder are disjoint. Subsequently, the ILP selects the optimal set of DFOS placements to minimize the total number of deployed sensors while meeting critical infrastructure monitoring requirements, such as redundant monitoring for high-importance links. The method enhances the observability of critical links, achieving 100% monitoring coverage for important links during simulated power outages affecting an electric feeder.

A Framework for developing, training and deploying Predictive Analytics Models using Artificial Intelligence based Machine Learning algorithms is presented. The Framework is a collection of processes with User Interface and Data Processing components which interact with input data acquisition, output data visualization and data transmission systems. The key part of the Framework is automation of processes involved in building, training and deploying a model, with the aim of reducing the time and manual effort normally required in such tasks. The Framework also includes a proprietary method of machine learning algorithm selection process based on use case and certain characteristics of data, that ultimately results in improved accuracy of Predictive Analytics models for real time and non-real time applications.

A method for proof-of-authority for a highly reliable blockchain network according to the present invention comprises the steps in which: a proof-of-authority consensus processing unit selects an extra signer through a preset priority algorithm in a case in which a malfunctioning signer is present when the proof-of-authority consensus processing unit monitors network activities in a blockchain network in which a plurality of approved signers verify a transaction and then include the approved transaction in a chain as a block when the transaction starts at a node; when the new extra signer is approved by majority approvals of the existing signers, the proof-of-authority consensus processing unit replaces the malfunctioning signer with the extra signer; and the new extra signer generates a block or verifies the generated block.