140 results about "Real time learning" patented technology

The invention provides a self-adaptive brightness adjustment method for an LED backlight source, and the method comprises the steps: obtaining the intensity and spectral distribution information of downward natural light and an indoor lamp light source in real time through a light sensor, carrying out the light source superposition, calculating the total light intensity and spectral distribution, judging whether the self-adaptive adjustment condition is met or not, and if yes, adjusting the brightness of the LED backlight source. If yes, triggering an LED backlight brightness adjusting process; detecting the actual brightness of each to-be-adjusted unit in real time, comparing the actual brightness with the optimized target brightness value, calculating a brightness error, introducing a machine learning algorithm to learn and predict error data in real time, and dynamically adjusting the PWM duty ratio of the to-be-adjusted unit; and continuously obtaining evaluation feedback data of the user on the display image quality, and performing clustering analysis on the evaluation feedback data through an optimized unsupervised learning algorithm to generate a brightness adjustment curve adapted to different user preferences.

The invention discloses an intelligent evaluation method for real-time learning and teaching effect feedback, and the method comprises the steps: carrying out the preprocessing of collected multi-modal teaching scene data of a target classroom, and obtaining a multi-modal data stream which comprises the aligned target video data, target audio data and a timestamp; pre-training the visual basic model based on the ViT architecture by adopting a plurality of education scene data sets to obtain a first model, carrying out alignment fusion on the first model and a preset variational automatic encoder structure to obtain a preset target model, and carrying out dimension reduction processing on the extracted spatial-temporal features and feature vectors by utilizing the model to obtain target features; and determining a classroom atmosphere index, a student participation index and a teaching interaction effect index based on teaching behavior characteristics, student behavior characteristics and teaching interaction behavior characteristics identified from the target characteristics, and synchronously generating a real-time teaching quality evaluation report. According to the method, the classroom information can be accurately and comprehensively analyzed, and a real-time and objective evaluation result is synchronously generated.

An AI-driven financial planning system for real-time market adjustment, consisting of: a neural inference coprocessor configured to execute deep financial forecasting models, including recurrent neural networks and attention-based encoders, on the device, and wherein the processor dynamically updates portfolio parameters in response to market signals exhibiting volatility differences above a statistical threshold calculated using an exponentially weighted moving standard deviation; a financial data acquisition module configured to continuously receive and analyze heterogeneous data streams, including market indices, interest rates, stock and bond price fluctuations, economic indicators, regulatory updates, and financial news sentiment feeds; a behavioral analytics engine configured to create a dynamically evolving user-specific financial behavior profile based on real-time analysis of transaction history, income-expenditure cycles, psychometric test results, and temporal lifestyle patterns using supervised and unsupervised machine learning algorithms; A goal optimization module configured to transform high-level, user-defined financial goals into quantitatively tracked multi-level goals. It uses a reinforcement learning framework that predicts optimal asset allocations across multiple time horizons. a real-time strategy simulation engine configured to perform Monte Carlo simulations and deep Q-learning-based assessments to simulate the resilience of proposed financial strategies under different macroeconomic regimes and trigger redistribution events based on predefined confidence thresholds; a compliance-aware execution interface configured to interact with financial institutions through encrypted API channels, ensuring policy enforcement using a smart contract validator and a hardware-enabled secure transaction signing unit; and a recommendation display unit configured to render dynamic dashboards for visualizing investments, reallocation warnings, confidence intervals, and sensitivity sliders, and where user interaction with the unit flows back into the behavioral model for real-time learning.

The invention provides a wisdom education method and system based on student behavior sequence recommendation, and relates to the technical field of wisdom education. According to the method and the system, through an AI autonomous identification knowledge graph construction and innovative knowledge fusion method, a course knowledge system is structurally and hierarchically displayed, and a knowledge framework is clearly presented, so that knowledge graph framework construction of all-platform courses is facilitated, pre-repaired courses and subsequent courses are efficiently associated, and the construction efficiency is improved. And a clear and complete course learning path is provided for students. By adopting a recommendation algorithm based on a student behavior sequence, combining SASRec and introducing a self-attention mechanism of a sparse attention module, behavior data of students can be deeply analyzed, a learning path is adjusted according to real-time learning data, and personalized learning suggestions are provided for the students. According to the method and the system, multi-dimensional functions such as knowledge graph construction, a student behavior sequence recommendation algorithm and intelligent learning path optimization are integrated, and the method and the system have wide coverage and strong comprehensiveness.

The invention relates to the technical field of computers, in particular to a personalized learning recommendation method and device based on AI workflow and electronic equipment, after a user authorizes, learning behavior data of the user on a teaching platform is collected in real time, based on the collected data, the current learning state and ability level of the user are evaluated, the learning requirement is determined, and according to the learning requirement, the learning state and ability level of the user are recommended. According to the method, the difficulty encountered when a user learns course content is predicted, a course set suitable for current learning of the user is screened out from a preset course library according to a prediction result and learning requirements, and a personalized course recommendation list is generated in combination with interest preference and real-time learning behavior data of the user. Meanwhile, by collecting and analyzing feedback suggestions of the user on the recommended course and adjusting the recommendation strategy in real time according to the feedback, immediate feedback and recommendation adjustment on new questions or demands of the user are achieved.

The invention provides a personalized English education system and method based on multi-modal sentiment analysis. The system comprises a multi-modal interaction module for receiving and processing multi-modal data of students, an emotion recognition module for performing emotion analysis on the input multi-modal data, and a personalized module for evaluating real-time learning states of the students based on historical learning data of the students, and the core brain module is used for dynamically adjusting interactive feedback according to the emotional state and the real-time learning state. The emotion recognition module comprises emotion information fusion, the emotion information fusion adopts a weighting strategy, and the final output emotion state is adjusted through emotion consistency constraint and a conflict correction mechanism. According to the invention, through an emotion consistency loss function, a conflict correction mechanism and a knowledge graph-based super-outline control mechanism, the emotion and cognitive states of the students are accurately identified.

The invention provides a collaborative decision-making system based on multi-agent reinforcement learning, and the system comprises a global decision-making coordination module, an agent local decision-making module, a state information sharing module, a conflict detection and solution module, and a security and privacy protection module. Each agent realizes autonomous decision making and real-time learning in a local environment, the dependence of the system on centralized control is reduced, the communication burden of the system during data interaction is effectively reduced in a distributed manner, the agents can directly respond based on local data, and the data interaction efficiency is improved. The state information of each agent is collected and analyzed in real time through a global decision coordination module, and the overall strategy of the system is adjusted based on a reinforcement learning algorithm, so that global target implementation in a complex environment is ensured, and the limitation that a traditional system cannot be adaptively adjusted in a dynamic environment is effectively avoided through global optimization; and the global optimization of the multi-agent cooperative system on the basis of each agent target is ensured.

The invention discloses a teaching display system and method based on a knowledge graph, and a medium, and relates to the technical field of artificial intelligence and education, and the method comprises the steps: regularly collecting multi-modal original teaching data, and constructing a teaching knowledge graph; dynamically updating the relation weight of the teaching knowledge graph; according to the potential capability coefficient of each learning terminal, the distinction degree coefficient and the difficulty coefficient of the plurality of questions corresponding to each knowledge point and the relationship weight between the knowledge points, obtaining a cognitive dependency coefficient between the knowledge points corresponding to each learning terminal; generating a personalized learning path of each learning terminal according to the cognitive dependency coefficient and the learning target; performing interaction attribute addition on nodes in the teaching knowledge graph according to the personalized learning path, and generating a VR virtual scene and visual data; and real-time learning monitoring is performed on the personalized learning path of each learning terminal, and real-time learning feedback operation is performed according to the real-time learning monitoring result, so that the overall teaching quality is remarkably improved.

The invention relates to an electrical equipment fault monitoring and positioning method and system based on data analysis, and belongs to the technical field of electrical equipment monitoring. The method comprises the following steps: sensing a transient traveling wave signal of a fault current by adopting a traveling wave detection technology to determine an initial fault area; acquiring state data of the secondary equipment in the fault area through mapping of the secondary equipment in the fault area; constructing a secondary loop connection model based on the configuration file, and designing an action sequence rule base to verify the action logic and time sequence matching of a regional data centralized protection device, a circuit breaker and a communication link in real time; based on logic verification, action triggering conditions are extracted to be fused with the regional data set; and inputting the fault feature vector into a classification model for fault identification by obtaining the classification model. According to the invention, through traveling wave detection, EEMD decomposition, data synchronization and logic verification, feature information related to the fault is accurately extracted, and through data fusion and real-time learning, the fault identification precision and efficiency are improved.

The invention discloses a self-learning intelligent question answering system and method with long and short term memory, and the system comprises a self-learning module, a large model reply module, a short term memory module, a long term memory module, and a safety guarantee module. The self-learning module realizes real-time learning and updating, so that the system can be continuously improved according to latest user interaction; the long-short term memory module processes short-term information and long-term information respectively, so that the system can flexibly meet various different types of user requirements, and accurate and timely answers are provided. According to the technical scheme, the technical problem that an existing system lacks self-learning ability and long-short-term memory design is solved, the intelligent level and user experience of the system are remarkably improved, more intelligent and efficient services are provided for a user, and the advantages and the positive effects are beneficial to popularization and application. Therefore, the system has strong competitiveness and wide application prospect in the market.

The invention relates to the technical field of data center interaction, and particularly discloses an information interaction method and system for computing power and electric power collaborative scheduling of a data center, which integrates multi-source heterogeneous data and constructs a space-time correlation model to realize accurate prediction and dynamic modeling of resource demands of the data center. The real-time learning ability of the DRL and the discrete decision advantage of the MIP are utilized to realize multi-objective synchronous optimization, the energy efficiency and economy of the data center are significantly improved, the multi-objective optimization can realize "computing power-electric power-heating power" coupling, the block chain technology ensures that data cannot be tampered and is transparent and credible, the security and auditing performance of the system are improved, and the system performance is improved. Carbon footprint tracking, green power authentication and PUE index dynamic energy efficiency optimization are supported, green power use is directly stimulated, the carbon footprint and operation cost is reduced, thermal data is used for cooling control, power data is used for task scheduling, resource conflicts are avoided, the limitation of traditional single resource scheduling is broken through, and the response speed is increased.

The invention discloses a heat supply load prediction system, and relates to the technical field of heat supply prediction. The data acquisition and fusion module is used for acquiring multi-source heterogeneous data in real time and integrating the acquired multi-source heterogeneous data into a data set with a uniform format; the feature engineering module extracts key features influencing the heat supply load, and a feature subset with the highest prediction value is screened out through a feature selection method and serves as input of a deep learning model; constructing a deep long-short-term memory network model based on an attention mechanism, and performing model training through the training data set; the real-time prediction and adaptive adjustment module receives new data acquired by the data acquisition and fusion module in real time, and inputs the new data into the trained deep long and short term memory network model based on the attention mechanism for real-time prediction of the heat supply load; by integrating multi-source heterogeneous data and utilizing an advanced deep learning model, accurate prediction of the heat supply load is realized, and the method has real-time learning and adaptive adjustment capabilities.

The invention relates to the field of resource optimization, and discloses a supply chain full-link-oriented logistics resource optimization and matching method and system, and the method comprises the steps: generating a multi-dimensional logistics state feature vector according to full-link multi-source data of a target supply chain, constructing a multi-target optimization model according to a fuzzy inference rule and a fuzzy weight system, and carrying out the optimization of the multi-target logistics state feature vector; performing preliminary matching degree analysis on the multi-dimensional logistics state feature vector, performing resource allocation on a target supply chain by using a preliminary logistics resource allocation scheme generated according to a comprehensive matching score, and performing weight online learning and dynamic self-adaptive adjustment on a fuzzy weight system according to a performance error value of an actual performance parameter value, so as to obtain a multi-dimensional logistics resource allocation scheme; and obtaining a target matching weight system, optimizing the preliminary logistics resource allocation scheme, and matching the logistics resources of the target supply chain to obtain a target resource matching result. According to the invention, in a dynamic scene of a full link of a supply chain, real-time learning of a multi-target tradeoff relation and adaptive updating of a fuzzy weight can be realized.

The invention relates to the technical field of data processing, in particular to a learning behavior data analysis system and method based on an agent, and the method comprises the steps: determining knowledge points in learning content, generating a visual knowledge point graph, testing the mastering degree of a learner for the knowledge points, generating a visual knowledge point mastering graph, and analyzing the knowledge points. The method comprises the steps of generating an optimal learning task which is most adaptive to the state of a learner based on the mastery degree of the learner and a learning behavior library, collecting real-time learning feature data of the learner during task execution, sensing the emotion and attention state of the learner in real time, and generating a personalized learning path in combination with a knowledge point graph, learning behavior analysis and the state of the learner. Learning tasks are dynamically adjusted according to real-time data, and learning strategies are optimized according to feedback of learners. Through dynamic perception and adjustment capability, the problem that a learning behavior analysis system in the prior art lacks real-time perception and dynamic adjustment of the state of a learner can be effectively solved.

The invention provides a multi-modal interaction and image processing method for AI glasses, electronic equipment and a storage medium, and the method comprises the steps: carrying out the preprocessing of an original image collected by the AI glasses in real time based on a space-time attention mechanism, and obtaining a preprocessed image; performing voice recognition and semantic analysis on a voice instruction input by a user through the AI glasses to obtain a text instruction; performing keyword matching on the text instruction and a preset image processing instruction template library, and determining a corresponding image processing strategy; if the image processing strategy is dynamic tracking, predicting the position of a target in the preprocessed image in the next frame by using a Kalman filtering algorithm, and searching and positioning the target around the predicted position by using a twin network; and the appearance features of the target are learned and updated in real time based on the twin network, stable tracking of the target is realized, and a tracking target image sequence is output. According to the method, the defects that the subsequent appearance position of the target cannot be predicted and the target cannot be stably tracked at present are overcome.

The invention discloses a pollution source monitoring method and system based on a neural network, and relates to the technical field of pollution source monitoring, and the method comprises the steps: dividing a preset range of a target drinking water source into a plurality of grid sub-regions through a GIS means, setting water quality monitoring point positions in the grid sub-regions, and obtaining the water quality monitoring data of each point position; constructing a digital twinborn model, judging whether a water quality abnormity alarm signal and a water quality abnormity early warning signal are generated or not according to the water quality monitoring result of each water quality monitoring point, and obtaining a correlation coefficient between the water quality abnormity early warning signal generated by each water quality monitoring point and the scene sub-sequence; adding a pollution source monitoring model through a neural network on the basis of the digital twin model; according to the method, the pollution source monitoring model is learned in real time, real-time monitoring and control are performed by using the sewage treatment whole-process model according to the prediction result of the pollution source monitoring model, effective early warning of sudden water source water quality pollution events such as natural disasters in the drinking water source is realized, and the pollution monitoring timeliness is ensured.

The invention provides an artistic course personalized learning path method and system based on deep learning. Firstly, an artistic course learning basic information set of a learner and a preset artistic course module library are acquired; calling a pre-trained deep learning course dynamic association model to generate association strength description of the learning basis of the learner and each course module; screening and sorting based on association strength description to form an initial learning module sequence; acquiring a real-time learning behavior data set of the first course module learned by the learner, and generating a module adjustment signal; and adjusting the initial learning module sequence according to the module adjustment signal to obtain a personalized learning path adaptive to the real-time learning state of the learner, and generating a personalized learning path document comprising a course module learning sequence and module connection guidance, thereby realizing accurate customization and dynamic optimization of the learning path, and improving the learning efficiency. And the learning effect of artistic courses is improved.

The invention relates to the technical field of intelligent evaluation, and particularly discloses an intelligent evaluation model dynamic optimization method based on reinforcement learning, which is characterized in that a basic intelligent evaluation model and general parameters thereof are finely adjusted according to historical interaction data of a student object, and current interaction data of the student object is predicted by updated model parameters. A reference prediction error between a reference prediction result and real interaction is calculated, recent performance statistical characteristics and current personalized adjustment parameters of a student object are integrated as a current reinforcement learning (RL) state, an RL Agent strategy network is used for selecting an adjustment action, and the personalized adjustment parameters of the student object are calculated and updated based on the adjustment action. According to the method, personalized adjustment is performed on the real-time learning state of each student individual, so that the stability and generalization ability of the basic model are ensured, and quick response and targeted adjustment on the state change of the student individual are realized.

The invention provides a power grid standard intelligent recommendation method based on reinforcement learning and post system management. The power grid standard intelligent recommendation method comprises the following steps of 1, collecting and cleaning power grid standard data; step 2, post-standard system knowledge graph construction; 3, designing a reward function for the power grid standard recommendation system, and realizing a personalized recommendation strategy by taking post responsibilities and skill levels as dynamic weight parameters; and step 4, post-driven online incremental training is carried out. According to the method, a reinforcement learning dynamic optimization engine is adopted to learn user behaviors (click rate and task completion score) in real time, and a standard recommendation strategy is adjusted; and when the standard is updated, the post capability weight is automatically adjusted, closed-loop feedback optimization is realized, and the problem of updating delay of a traditional system is solved.

The invention relates to the technical field of machine learning, and discloses a compressor vibration adaptive suppression method and device based on reinforcement learning, and the method comprises the steps: obtaining a vibration quantity quantization parameter; obtaining a vibration trend quantization parameter based on the vibration quantity quantization parameter; constructing a reinforcement learning model based on the vibration quantity quantization parameter and the vibration trend quantization parameter; training a reinforcement learning model; deploying a reinforcement learning model, and generating a compensation current signal # imgabs0 # according to a compensation torque current parameter I and a compensation angle parameter theta output by the reinforcement learning model; and the frequency converter receives the compensation current signal # imgabs 1 #, superposes the compensation current signal # imgabs 2 # and a compressor motor current control signal, and drives the compressor motor to operate by using the superposed current control signal so as to suppress vibration caused by fluctuation of pressure and flow experienced by the compressed gas in each working cycle. According to the invention, through real-time learning and optimization, accurate suppression can be realized without a sensor.

The invention provides a target tracking detection method of a photoelectric pod, detection equipment, a program product and a storage medium. Tracking is carried out while target detection is carried out, and the delay time between detection and tracking is shortened. After the target detection result and the target tracking result are obtained, the positive sample and the negative sample are obtained according to the target detection result and the target tracking result and the video image, and in this way, target and non-target data can be accurately collected, so that training samples are more targeted and diversified. And training a target detection updating model by using positive and negative samples, so that the model can learn the change characteristics of the target in real time, and outputting more accurate characteristic data and a target model for detecting and tracking a new video image. And finally, a correct target is determined through confidence score, and the tracking detection precision of the autonomous or semi-autonomous aircraft on the target in a complex and changeable environment is improved by integrating the interaction of the technical characteristics.

The invention discloses an electro-hydraulic system adaptive neural network preset time control method, which integrates a neural network real-time learning technology, innovatively introduces a preset time performance function, and designs an adaptive neural network preset time performance controller. For the electro-hydraulic system position tracking control problem, it can be ensured that system output transient performance and steady-state performance converge to a specified performance index range within a preset time, safe and reliable operation of the system is ensured, the unknown dynamic state of the system can be learned in real time through a neural network, high-precision motion control performance is achieved, and the control precision is improved. And the problem of differential explosion in traditional backstepping control of the electro-hydraulic system can be avoided, and the influence of measurement noise on the control precision is reduced.

The invention discloses an AIGC enabled Chinese and foreign culture comparative auxiliary teaching system and teaching method. The system dynamically generates texts, images and audio and video data closely related to foreign culture by means of the AIGC technology according to user requirements and learning levels. Meanwhile, the system can fully consider the difference of education ideas and learning habits under different cultures, formulate a personalized learning scheme for each user, and provide real-time learning effect evaluation and feedback. According to the method, deep learning, natural language processing and multimedia display technologies are applied, and the functions of user interaction, multi-modal learning, resource updating management, system optimization and upgrading and the like are integrated. The language skills of the user are improved; and foreign culture connotation is deeply explored. On the level of Chinese and foreign culture comparison, comprehensive, interactive and personalized culture learning experience is provided for users, the method is suitable for multiple culture teaching scenes, communication and understanding of Chinese and foreign culture are promoted, learners are helped to cross culture differences, and more effective cross-culture communication is achieved.

The present invention discloses an optimization method for dynamically generating a learning path of a large language model based on RAG, which relates to the technical field of data processing and includes: obtaining a predetermined learning path in a predetermined learning syllabus; obtaining an RAG model; performing information retrieval and analysis on a target dynamic learning portrait to obtain a target real-time learning state, and performing a correlation comparison analysis with a predetermined knowledge base to obtain a candidate knowledge sequence; performing knowledge fusion analysis on the target predetermined knowledge and the candidate knowledge sequence to obtain a target knowledge sequence, and dynamically adjusting the predetermined learning path in combination with a target period to generate a target dynamic learning path for a target student. The present invention solves the technical problem that the prior art cannot perform intelligent dynamic adjustment of the subsequent learning path according to the real-time learning situation of students, resulting in a lack of pertinence in the learning process of students, unsatisfactory learning quality and effect, and low efficiency, and achieves the technical effect of improving the personalization and flexibility of the learning path and enhancing the overall learning effect of students.

The invention relates to the technical field of large language models, in particular to an AI large model course recommendation and adaptive learning path optimization system for multi-dimensional learning condition analysis. Comprising a learning condition data processing unit; a multi-dimensional learning condition analysis unit; an AI large model course recommendation unit; and an adaptive learning path optimization unit. According to the method, through a matching degree calculation module and a parameter optimization module of an AI large model course recommendation unit, cosine similarity is calculated based on course joint features and learning condition features, dynamic matching degrees are obtained in combination with course-geographical attention weights, courses with the dynamic matching degrees larger than a threshold value are screened, and the first N courses are taken in a descending order to generate a personalized recommendation list; and then real-time learning effect data when the user learns the recommendation list is received, the model parameters are iteratively updated by adopting an Adam optimizer after the optimization loss is constructed, and the problem of lack of iterative optimization of the model parameters based on the real-time learning effect data in the prior art is solved.

The invention relates to the technical field of intelligent education planning, and discloses a dynamic course planning method based on a knowledge graph. The method comprises the following steps: acquiring real-time learning data including knowledge point mastering degree, learning behavior and progress; and performing association analysis on the data by using the subject knowledge graph to generate a learning state representation vector reflecting knowledge point association strength and path dependency degree. And inputting the vector into a course planning model for calculation to obtain an adjustment scheme containing the knowledge points to be strengthened, the path rearrangement sequence and the resource index. According to the scheme, the node weight of the corresponding learner in the knowledge graph is updated, and a dynamically evolved personalized knowledge graph is formed. And generating a next-stage learning task instruction based on the graph, wherein a logic mapping relationship between the to-be-learned knowledge point sequence and a preset learning target is clearly revealed. According to the method, structured traceability and diagnosis of learning weak links are realized, and an interpretable personalized learning path with a clear logic basis can be provided.

The invention discloses an online learning evaluation method and system based on artificial intelligence, and relates to the technical field of learning evaluation, and the method comprises the steps: collecting and preprocessing the learning data of employees, analyzing the attention of the employees to each knowledge point through a ruminant ratio algorithm, and generating a learning behavior feature matrix; and generating a preliminary learning path based on the learning behavior feature matrix, and dynamically adjusting the learning path according to the real-time learning behavior of the employee. According to the method, by using a ruminant algorithm and combining an artificial intelligence technology, not only can the attention degree of employees to each knowledge point be accurately analyzed, but also important support can be provided for subsequent path optimization and personalized recommendation by generating a learning behavior feature matrix; through combination of the graph neural network and the dynamic time warping algorithm and calculation of potential capability representation and demand variables of the nodes, deep analysis and dynamic adjustment of employee learning behaviors are realized in learning path generation, and individual differences of employees can be better adapted.

The invention provides a departure training recommendation method and device based on a grouping model, and the method comprises the steps: a departure user behavior data collection step: collecting various types of data when a user uses a departure system, and carrying out the analysis and storage of a plurality of dimensions; a departure user grouping step: according to the collected data, performing grouping by using a clustering algorithm to obtain different user group type function lists; a training content recommendation step: real-time learning progress analysis is carried out, and a personalized training recommendation list is recommended for the user; and a departure front-end display step: displaying the training function recommendation list to the user, constructing test data based on the service object characteristics, and customizing a personalized training scheme for the user. According to the method, the clustering algorithm is adopted, the operation behavior data of the users are deeply mined and analyzed, the requirements of each user can be more accurately recognized, and personalized recommendation is achieved; and an advanced stream processing technology and a real-time analysis framework are utilized to respond to a user request in real time.

The invention discloses a large language model generative recommendation method based on self-adaptive reflection mechanism driving, and relates to the technical field of recommendation. The method comprises the steps of performing online real-time learning on behavior data of a user, and determining an initial recommendation result through a large language model; performing automatic feedback learning through the initial recommendation result, and calculating a three-layer reflection score and a comprehensive reflection score; according to the three-layer reflection score and the comprehensive reflection score, executing total regeneration or local regeneration on the initial recommendation network, and regenerating an initial recommendation result; the newly generated initial recommendation result is evaluated again, a new comprehensive reflection score is determined, and if the new comprehensive reflection score is smaller than a comprehensive score threshold value, local fragments of the initial recommendation result are optimized, and a local generation result is obtained; and carrying out weighted fusion on the local generation result and the newly generated initial recommendation result to obtain a global recommendation result. According to the method, the recommendation efficiency can be improved while the recommendation precision is ensured.

The invention provides a dynamic intelligent target detection system and method based on a reasoning, namely a training normal form, and the system comprises a real-time learning module which carries out the real-time reasoning of input data through a YOLOX model, monitors a model reasoning result in real time, and triggers and activates the online learning according to the needs; the parameter fusion engine is used for training a model needing online learning, calculating parameter sensitivity based on Hessian matrix approximation by applying a gradient importance sorting algorithm, filtering abnormal node gradients through a difference perception aggregation protocol and by using cosine similarity, and realizing stable parameter transition through momentum updating by adopting progressive knowledge fusion; the multi-stage verification subsystem is used for carrying out automatic inspection on the trained model; and the resource awareness controller is used for dynamically adjusting resource allocation and gradient sparse transmission by acquiring hardware resource conditions in real time. The real-time performance, the accuracy and the environment adaptability of target detection can be remarkably improved, and it is ensured that the model efficiently runs on different hardware platforms.