1070 results about "Decision tree model" patented technology

The invention provides an intelligent short message scheduling method and device based on multi-dimensional dynamic optimization, and the method comprises the steps: obtaining the performance data of a plurality of short message channels, and calculating a channel health score based on a weight dynamic adjustment model; determining a scheduling strategy according to the priority identifier of the to-be-sent message, and performing channel screening and optimal matching; executing message sending and monitoring a sending state; terminal state detection is carried out on the failure message through operator base station signaling, and a decision tree model is applied to determine a retry strategy; and performing Huffman coding compression processing on the P2-level marketing messages which fail in retry, and performing batch sending in an idle window. According to the method, a comprehensive performance evaluation index and reward function model is also constructed, and parameter optimization is performed by applying a reinforcement learning algorithm. According to the invention, multi-dimensional dynamic channel scoring, intelligent retry decision making based on terminal state perception, batch processing with balanced cost-time efficiency and a closed-loop self-optimization system are realized, the short message delivery rate is obviously improved, and the invalid retry rate and the sending cost are reduced.

The invention discloses a bolt on-line monitoring method and system, and the method comprises the steps: generating a micro-motion feature matrix of a bolt according to a multi-dimensional vibration signal and a stress wave propagation characteristic of a bolt connection part; based on the micro-motion characteristic matrix, fusing local stress distribution data acquired by a bolt surface strain gauge, and outputting a health state quantitative index of the bolt; outputting a loosening risk level according to the health state quantitative index and historical loosening evolution data; and based on the looseness risk level and in combination with equipment operation state parameters, a self-adaptive decision tree model is adopted to dynamically adjust monitoring frequency and an alarm threshold value, a real-time monitoring strategy and a visual early warning report are generated, and a wireless transmission module is synchronously triggered to upload the real-time monitoring strategy and the visual early warning report to a cloud operation and maintenance platform. According to the embodiment of the invention, high-precision, self-adaptive and traceable bolt health state evaluation and early warning can be realized.

The invention discloses a time-space alignment fusion processing method and system for multi-source physiological signals, and relates to the technical field of data processing.The method comprises the steps that the multi-source physiological signals in the limb movement state are synchronously collected, and a heterogeneous time sequence data set is obtained; performing space-time alignment processing on the heterogeneous time sequence data set to generate a synchronous physiological signal matrix; performing signal quality evaluation based on the synchronous physiological signal matrix, constructing a weighted decision tree model, performing confidence fusion on the synchronous physiological signal matrix according to the weighted decision tree model, and generating multi-parameter joint monitoring data; and performing motion artifact suppression processing on the multi-parameter joint monitoring data, outputting a physiological parameter index set, and transmitting the physiological parameter index set to a first-aid equipment monitoring terminal. Therefore, the technical effects of eliminating signal distortion, improving monitoring data quality and ensuring first-aid monitoring precision are achieved.

The invention provides a combined wind power prediction method suitable for a distributed wind power plant, and the method comprises the steps: collecting the real-time meteorological data and historical power data of a wind power plant cluster, carrying out the cross-wind-plant data collaborative cleaning, and generating a time-space aligned standardized data set. Constructing an adaptive spatio-temporal feature extractor, outputting a spatio-temporal feature matrix, and inputting the spatio-temporal feature matrix into the spatio-temporal adaptive neural network, the graph attention prediction model and the physical constraint decision tree model to generate three prediction sequences. And the sequences are fused through a space-time collaborative attention mechanism to generate a dynamic weighted combination prediction result. And performing physical constraint correction on the result by using a space-time residual error correction network to generate a final prediction sequence. And updating the neural network topological structure based on the prediction error distribution, and outputting a prediction result with uncertainty evaluation to a power grid dispatching system. According to the method, the precision and reliability of wind power prediction of the distributed wind power plant can be improved, and the stability and economy of power grid dispatching are improved.

The invention relates to the technical field of data processing, in particular to a backtracking analysis model construction method based on an attack chain, which comprises the following steps that: a kernel layer security agent acquires process, file and network behavior characteristics in a hardware isolation environment, and generates an event tuple; the tensor network pipeline performs three-dimensional decoupling mapping on the tuple into a behavior fingerprint vector, an orthogonalization noise feature and an asymmetric adjacent tensor, and compresses the behavior fingerprint vector, the orthogonalization noise feature and the asymmetric adjacent tensor into a space-time topology tensor block; the reinforcement learning controller constructs a directed acyclic graph based on the tensor blocks, calculates connectivity loss and outputs an event risk score; the dynamic routing engine constructs a decision tree model according to the risk mark, the burst frequency and the correlation entropy, and implements three-level shunting and a multiple simulation system to generate an anti-interference index; and when the deviation between the physical trajectory and the digital model exceeds the tolerance, the closed-loop feedback weight coefficient updates the loss function parameter and adjusts the channel resource weight. And the problem of threat discovery delay caused by attack chain breakage under massive events is solved.

The invention discloses a P4 switch processor core-oriented multi-service traffic packet scheduling method, which mainly comprises the following steps of: proposing a feature extraction method based on a statistical interval and Sketch, integrating data packet time correlation and traffic dynamic change, optimizing a Bloom filter conflict detection mechanism, and improving the adaptability and accuracy of an algorithm in a limited resource environment; a decision tree model is combined with scheduling, a decision tree is reconstructed to adapt to an operation mode and a data processing requirement of a P4 switch, and high-precision service flow classification under linear speed operation is ensured; a packet scheduling method AD-PIFO is designed based on a variable priority, the forwarding priority of a data packet is dynamically adjusted, the QoS requirements of different services are met, and the overall performance and efficiency of the switch are improved. The method is outstanding in high-priority traffic throughput and low-delay performance, is superior to similar algorithms in medium-low priority traffic throughput, is stable and effective under different load conditions, is reasonable in extra time overhead of data packet transmission, and is widely applied to SDN network optimization data packet scheduling scenes.

The invention discloses a monitoring and early warning method and system based on active load prediction of a power distribution network, which breaks through the limitation of purely depending on historical data and simple statistical analysis, fully considers the complexity and uncertainty of load change influenced by various factors, and achieves the early warning of the load change through the combination of multiple models. The method can accurately classify and analyze the operation active load of the power distribution network through an active power analysis model, a decision tree model and the like, thereby improving the prediction precision. In the aspect of data processing, mass operation data generated by intelligent equipment in the power distribution network can be effectively integrated and mined by utilizing a vector machine model and a cloud large model server, an early warning mechanism is flexible, the method can adapt to different operation scenes and load characteristics, false alarm and missing alarm are avoided, and the method is suitable for popularization and application. According to the method, the power distribution network fault identification efficiency of the power distribution network power factors in different time periods is remarkably improved, and effective operation of the power distribution network is ensured.

The invention discloses a risk level prediction method and device for a power distribution network, and electronic equipment, and relates to the technical field of power distribution network operation and maintenance or other related fields, the method comprises the following steps: obtaining current operation data of a target power distribution network, the current operation data being acquired by a sensor network in real time, comprising line equipment operation data of the target power distribution network, geographical environment data of the position where the target power distribution network is located and disaster weather data; all the current operation data are input to a target model, the risk prediction level and early warning information are output, the target model is a decision tree model pre-established based on a random forest algorithm and is used for extracting key features from the line equipment operation data, the geographical environment data and the disaster weather data, and therefore risk level study and judgment are carried out; and early warning information is generated according to different risk prediction levels. According to the invention, the technical problem of low power distribution network risk prediction accuracy in extreme weather in the prior art is solved.

An application system fault prediction method based on an association rule and a deep learning integrated model, the method comprising: collecting historical fault data of an application system; performing data preprocessing on the collected historical fault data to obtain a data matrix, and then converting the data matrix into a data form suitable for analysis by an association rule algorithm and a deep learning algorithm; performing feature extraction on the data matrix to obtain effective fault-related features; establishing an association rule model and a deep learning LSTM model; fusing outputs of the association rule model and the deep learning LSTM model to form a multi-layer decision tree model; and predicting a fault by combining the association rule model, the deep learning LSTM model, the multi-layer decision tree model, and current state data of the application system, and obtaining a fault prediction result via weighted voting. The present invention achieves accurate fault prediction and early warning for an application system, and enhances the accuracy and robustness of fault prediction by introducing optimizations for a deep learning LSTM model.

An approach is provided for organizational transformation from a current state to a target state. Common language model(s) can dynamically perform interviews with stakeholders as part of a cooperative game to use disparate stakeholder insights to define the target state, projects, milestones, tasks, and resource use / availability. Lookalike Models can be used to model the organization as a dissipative system and calculate an organizational entropy score. A Markov model identifies possible task completion pathways between current and target state. An optimal project completion path through the Markov model may be identified using Decision Tree Models to identify magnitude of contribution to organizational transformation towards target state for each project and likelihood of successful project completion for each project using Fault Tree Models. Project completion resource allocation plans can be generated based on optimal Markov path. Bayesian Priors can be calculated based on performance measured using micro-behaviors analysis.

The invention provides an environment detection method and system based on a sample target detection model. The method comprises the following steps: synchronously acquiring an environment image, a video stream and physical parameters by using a multi-mode sensor; decomposing the data into image features and environmental parameter components through a dual-time sequence control signal, and realizing space-time alignment by adopting a linear phase filter; constructing a foreground region template based on the depth information, and generating target recognition feature representation containing an abnormal blurred target; adversarial training is carried out on the lightweight target detection network in combination with a transfer learning strategy, the network integrates convolutional features and a Transform attention mechanism, and the weight is dynamically adjusted through environmental parameters; fusing a target result and sensor data in real-time detection, and inputting a decision tree model for risk grading; and after the early warning is triggered, reconstructing a false detection sample through an online learning mechanism and iteratively optimizing the model. The system correspondingly comprises a multi-modal data acquisition module, a data enhancement and annotation module, a model training module, a real-time detection and fusion module and an early warning and optimization module. According to the invention, through multi-source data fusion, dynamic data enhancement and an adaptive compensation mechanism, the small target detection precision, the environmental adaptability and the real-time early warning capability are significantly improved.

The invention belongs to the technical field of optimization management, and particularly relates to an infrastructure site risk operation optimization management method based on an artificial intelligence algorithm, and the method comprises the steps: obtaining a historical infrastructure construction description set corresponding to each risk operation based on the construction monitoring record of each risk operation of a historical infrastructure site; performing data preprocessing to obtain a sample capital construction description set; performing classification based on the sample capital construction description set, obtaining a sample capital construction site feature set and a sample capital construction personnel feature set, and obtaining an intelligent decision tree model; and obtaining a real-time infrastructure construction description set corresponding to each risk operation, obtaining an intelligent risk decision value based on the real-time infrastructure construction description set and the intelligent decision tree model, and obtaining an infrastructure site risk operation scheme corresponding to each risk operation based on the intelligent risk decision value. According to the invention, accurate assessment and global coordinated optimization of capital construction site risk operation can be realized, and decision scientificity and management efficiency are improved.

The invention discloses a transaction card swiping early warning method and device based on AI identification, and relates to the technical field of financial payment, and the method comprises the steps: collecting transaction data, carrying out the preprocessing, and carrying out the feature extraction according to the preprocessed data, and generating a feature set; training an XGBoost model by using the generated feature set, and generating a basic classification result after constructing a preliminary classification model for transaction risk identification; the basic classification result is introduced into a decision tree model to carry out further optimization classification on the edge sample, and the obtained optimization classification result and the basic classification result are fused to generate a transaction risk score; according to the method, the multi-dimensional features of the nodes and the hyperedges are combined, the high-order relationship in the transaction data is comprehensively captured, more accurate features are constructed for expression, the node feature propagation path is optimized layer by layer based on the graph convolutional network and the dynamic factor adjustment mechanism, and the transaction efficiency is improved. And the robustness of the model to high-dimensional data and the analysis capability of the model to complex association are ensured, so that the early warning is more accurate and efficient.

A dynamic real estate valuation system based on multiparametric market indicators, which includes the following: a valuation engine configured to generate real-time results for property valuation; a multitude of distributed data ingestion and processing units configured to capture heterogeneous data sources, including historical property transaction data, real-time property listings, zoning and land use records, macroeconomic indicators, geospatial data, environmental sensor outputs, and sentiment-derived metrics; a model orchestration control unit comprising a stack of machine learning models, wherein the models include at least a gradient boosting decision tree model, a long-short-term memory (LSTM) time series forecaster, and an enhancement learning module that iteratively optimizes the model parameters based on the observed evaluation accuracy; a data contextualization controller configured to apply dynamic weighting to each input parameter based on the geographic, temporal, and market context by executing decay functions and location-specific rule matrices; a physical property valuation terminal (PVT) that includes an edge processing unit (EPU), geolocation circuitry, secure communication interfaces and a touch-based user interface; a valuation book subsystem configured to hash the valuation output, timestamp, and signatures of the input record into a blockchain-based distributed ledger; wherein the system is designed to continuously recalibrate its valuation results by comparing the predicted valuations with the actual sales or rental prices, and wherein the physical terminal is designed to produce a legally certifiable valuation document with embedded provenance data.

The invention relates to the technical field of automobile electronic testing, and discloses an integrated ECU power-on burning test monitoring system and method. The system comprises a power supply intelligent regulation and control module, a burning verification module, a test diagnosis module, an abnormity monitoring module and a feedback optimization module. The power supply intelligent regulation and control module realizes stable power supply output through a dynamic voltage frequency regulation algorithm and a temperature self-adaptive algorithm; the burning verification module is used for ensuring the burning integrity by using a block redundancy check and increment verification algorithm; the test diagnosis module diagnoses faults based on the multi-dimensional test data and the decision tree model; the abnormity monitoring module monitors abnormity in real time by using real-time streaming data processing and an isolated forest algorithm; and the feedback optimization module optimizes test diagnosis through incremental learning and a Bayesian optimization algorithm. According to the system and the method, comprehensive and accurate monitoring and optimization of the ECU are realized, and the performance and the reliability of the ECU are effectively improved.

The invention relates to the technical field of pipe gallery fire analysis, in particular to an underground pipe gallery fire feature correlation analysis system based on an improved Apriori algorithm, and aims to comprehensively collect underground pipe gallery data through multi-modal sensing equipment and pre-process the underground pipe gallery data into a standardized three-dimensional time sequence data set to provide a high-quality data basis for subsequent analysis; a multi-dimensional feature association matrix is constructed based on the three-dimensional time sequence data set, feature association strength is quantified, and a multi-dimensional feature association basis is provided for fire risk mining; a fire risk rule knowledge base is constructed through a space-time weighted support degree optimization Apriori algorithm, effective association rules are accurately extracted, and the accuracy and reliability of fire risk early warning are improved; based on the fire risk rule knowledge base, a decision tree model is established, association rule weights of branches of the decision tree are analyzed, fire risk grades are divided, a composite early warning instruction set is generated, different coping strategies can be adopted for risks of different grades, and the pertinence and effectiveness of emergency response are improved.

The invention discloses a power transmission line icing intelligent early warning method, system and device and a medium, and the method comprises the steps: obtaining icing state data and meteorological data, and dynamically adjusting the collection frequency and a dormancy strategy; performing data preprocessing and cleaning on the acquired data; extracting time-frequency features through wavelet packet transformation and a self-attention mechanism, and fusing the spatial dependency relationship and cross-modal interaction information of multiple monitoring points by using a graph neural network to obtain enhanced icing state characterization; performing icing risk prediction by adopting a gradient boosting decision tree model to obtain an icing risk prediction result; and analyzing an icing risk prediction result by using an interpretable tool, identifying a key factor which has the greatest influence on icing risk prediction, dynamically adjusting an early warning level according to the key factor, and generating an early warning and maintenance suggestion. Therefore, the monitoring real-time performance and the early warning timeliness are improved.

The invention provides a software code defect detection method and system based on program code feature fusion, and the method comprises the steps: carrying out the grammar and structure inspection of a code through a static analysis tool, and extracting defect features related to control flow chaos, exception handling and resource leakage; running a preset test case through a dynamic analysis tool, capturing an execution path and an abnormal behavior in a runtime state, and obtaining defect features which are not covered by static analysis; if the output result of the decision tree model is that defects exist, grouping comprehensive defect feature vectors by adopting a density-based clustering algorithm, determining the types and distribution ranges of the defects, and selecting the density clustering algorithm to process defect features in non-spherical distribution; and according to a clustering result, generating a defect repairing suggestion, combining code logic and execution path information, and determining a specific position and an influence range of the defect. According to the method, the stealth defect in the code can be effectively identified and positioned, the software quality and reliability are improved, and timely and accurate defect repairing guidance is provided for developers.

The invention discloses an unstructured storage hierarchical strategy optimization method based on machine learning, and relates to the technical field of data storage management, and the method comprises the steps: monitoring a file access event in real time, generating an access log, and extracting a multi-dimensional feature data set; utilizing the trained multi-dimensional decision tree model to distribute a corresponding storage hierarchy for the storage object to obtain a storage hierarchy decision; according to a storage level decision, distributing the storage objects to different storage layers, and carrying out resource configuration and storage operation; monitoring the access condition of the storage object in the new storage hierarchy in real time, and collecting file access performance, storage cost and response time to obtain feedback data; by monitoring the file access event in real time and extracting the multi-dimensional feature data set containing the basic attribute, the access behavior and the context information, efficient response of hotspot data and reasonable utilization of resources are ensured, and the overall performance utilization rate and the cost effectiveness of storage are remarkably improved.

The invention discloses a workshop task allocation method and system based on multiple dimensions, and relates to the technical field of production control management. Comprising the steps of obtaining order data and then performing preprocessing; based on the preprocessed order data, an advanced machine learning algorithm is adopted to construct a decision tree model, and the processing complexity of each order is identified; determining a task allocation scheme by using an optimization algorithm according to the identified order processing complexity and the current state of the production line; monitoring the actual operation condition of each production line in real time, starting a dynamic adjustment mechanism when the work load of a certain production line is unbalanced, and re-evaluating and adjusting task distribution; through intelligent task allocation and resource optimization, the production efficiency and customer satisfaction are improved, and meanwhile, dynamic adjustment and intelligent management of the production process are realized.

The invention relates to the technical field of optical fiber networks, in particular to an optical fiber network failure judgment method and device based on a feedback decision mechanism, and the method comprises the steps: constructing an optical fiber network, setting an intelligent optical fiber distribution device at an intermediate node of the optical fiber network, and setting a service probe at a tail end node of the optical fiber network; feedback information is obtained; performing multi-source feature extraction on the obtained feedback information, wherein the multi-source feature extraction comprises optical power data feature extraction, OTDR data feature extraction and service probe data feature extraction; and analyzing the extracted multi-source features by adopting a rule-based and model-based dual failure judgment mechanism to obtain a judgment result of the current network state. Therefore, multi-source information fed back by the optical fiber network is comprehensively analyzed by a rule-based and model-based dual failure judgment mechanism, the comprehensiveness, accuracy and efficiency of failure judgment are improved, corresponding repair measures are given according to a fault judgment result of the decision tree model, the response time is greatly shortened, and the network availability is improved.

The invention relates to the technical field of intelligent payment, and discloses a payment interface market dynamic matching and optimization method, which comprises the steps of obtaining multi-dimensional data, and constructing a user portrait based on the multi-dimensional data; based on a stream processing engine, aggregating the user behavior data, and updating a user portrait; identifying the user portrait based on the decision tree model, determining a transaction scene, constructing a differentiation strategy, and generating a payment mode recommendation list based on the differentiation strategy; inputting the user portrait into a pre-trained probability model for comparison, and updating a payment mode recommendation list based on a multi-target algorithm; determining a payment channel state based on the multi-dimensional data, obtaining a comprehensive score of a payment channel based on the payment channel state, and generating a candidate channel based on the comprehensive score; and obtaining an equipment health degree state, and determining an optimal payment channel based on the equipment health degree state. According to the invention, the operation fluency of a user in a complex use scene is improved.

The invention relates to the technical field of street lamps, and provides a solar street lamp fault self-inspection and cloud alarm system, which comprises a cloud control platform and a local module, and is characterized in that the local module comprises a wireless communication unit, a data processing unit with a built-in dynamic decision algorithm, a data acquisition unit and an execution unit; the method comprises the following steps: periodically collecting real-time operation data and real-time environment data of the solar street lamp; outputting a fault type code and an emergency disposal instruction through the fault classification decision tree model; sending alarm information to the cloud control platform; the execution unit executes the emergency disposal instruction. According to the method, the dynamic threshold value interval is generated through the dynamic threshold value adjustment algorithm, the influence of dynamic environment parameters such as illumination intensity and temperature on the running state of the solar street lamp is fully considered, misjudgment caused by fixed threshold value judgment is avoided, the system false alarm rate is greatly reduced, and unnecessary manual troubleshooting cost is reduced.

The invention relates to the technical field of intelligent equipment, in particular to an intelligent interactive method and system for pet emotion pacifying based on the Internet of Things, and the method comprises the steps: collecting the audio data, physiological data, environment data and owner feedback data of a pet, and carrying out the data cleaning of the collected data; performing feature extraction; combining a decision tree model, a recurrent neural network, a variant of the recurrent neural network, a Transform model and a graph neural network, constructing a hierarchical fusion model, introducing a transfer learning and reinforcement learning mechanism, designing a dynamic model fusion weight adjustment strategy, and judging pet emotion according to extracted features; different pacifying measures are adopted according to the types of the abnormal emotions of the pet judged by the hierarchical fusion model; and establishing a multi-index evaluation system. According to the invention, through the hierarchical fusion model of technology fusion, the pet emotion recognition precision and speed are improved.

The invention relates to the technical field of data encryption, and discloses a data encryption method, encryption equipment and a storage medium. In the method, an encryption device obtains to-be-encrypted data uploaded by a terminal, and extracts a plurality of data features from the to-be-encrypted data, the data features including a data sensitive feature, a business risk feature and a user behavior feature; performing quantitative scoring on the plurality of data features to obtain a plurality of feature scores, and calculating a total data score according to the plurality of feature scores; inputting the total data score into a preset trained decision tree model to determine the security level of the to-be-encrypted data; according to the security level, determining a target encryption algorithm from a preset corresponding relationship between the security level and the encryption algorithm; and encrypting the to-be-encrypted data according to the target encryption algorithm and the dynamically generated encryption key. By means of the method, the problems that differential protection is difficult to achieve according to data characteristics in a related encryption method, and security risks are caused by key management staticization are solved.

The invention provides a three-dimensional setting printing correction method and device, equipment and a storage medium, and the method comprises the steps: scanning a three-dimensional setting, and obtaining a three-dimensional point cloud model, color grid data and a multispectral texture mapping graph; based on the three-dimensional point cloud model and the color grid data, determining an offset defect labeling graph and a parallax offset matrix; performing two-dimensional mapping and global correlation calculation on the multispectral texture mapping graph to generate a texture anomaly thermodynamic graph and an LAB color space deviation vector; inputting the parallax offset matrix, the offset defect labeling diagram, the texture anomaly thermodynamic diagram and the LAB color space deviation vector into a preset decision tree model, and generating a root cause classification label and a printing quality score; and based on the root cause classification label and a preset historical printing parameter database, constructing a reinforcement learning model, taking the printing quality score as a reward function of the reinforcement learning model, and outputting a printing correction scheme.

The invention discloses a semiconductor probe card electrical parameter extraction method and system based on an MEMS technology, and relates to the technical field of semiconductor testing, the method comprises the following components: initialization setting: aligning an MEMS probe card to a target test area on a wafer to be tested, firmly fixing the probe card and the wafer together by using a mechanical clamping device, and setting the MEMS probe card and the wafer to be tested; by integrating the decision tree model and the genetic algorithm, accurate mapping and optimization of semiconductor wafer surface features and optimal probe contact parameters are achieved, the intelligent level of testing is improved, the test time is shortened, the test efficiency is improved, and the test cost is reduced. The self-adaptability and stability of the testing process are remarkably enhanced, real-time adjustment of probe contact parameters is ensured through a dynamic monitoring and closed-loop control mechanism, the testing error is effectively reduced, the testing precision is improved, the optimal contact parameter combination is found through an optimization algorithm, the testing period is further shortened, and the testing efficiency is improved.

The invention discloses a database malicious behavior detection and blocking method and system based on multi-dimensional features, and the method comprises the steps: collecting performance index data in real time, and sequentially carrying out the smoothing and normalization of the performance index data, and constructing a time sequence feature vector; performing analysis and feature extraction on the network flow data to obtain an abnormal connection mode feature vector; inputting the collected multi-dimensional feature vectors into an improved factorization machine, and performing feature crossing and dimension reduction processing by using an MFB pooling technology to obtain dimension-reduced feature vectors; inputting the dimension reduction feature vector into a dual-mode collaborative detection mechanism composed of an XGBoost decision tree model and a DRL model, and outputting a dynamic risk detection result; starting a hierarchical blocking strategy based on the risk level in the dynamic risk detection result; the method realizes closed-loop protection of feature extraction, risk rating and hierarchical blocking, has adaptive threshold adjustment and batch detection acceleration capabilities, and ensures continuity and high efficiency of database services.

Techniques described herein relate to training and executing machine learning models configured to determine injury probabilities based on vehicle accident data. In some cases, a decision tree model may be constructed including various branching criteria based on particular vehicle accident data and injury ground truth data, including leaf nodes storing corresponding injury probabilities. A model execution component may execute the trained models to determine the probability of potential injuries associated with vehicle accidents. Based on the potential injury probabilities, a vehicle accident analysis system may identify target computer systems and / or target processes to be initiated on the target systems, based on the injury probabilities. In some example, the model execution architecture may be implemented using an event-driven system and / or cloud-based data storage and services to receive, store, and process data events associated with individual vehicle accidents.

The invention discloses an automatic suction removal method, device and equipment for welding smoke and a medium, and relates to the technical field of machine learning. Linear interpolation and capacity expansion are conducted on a temperature matrix of a welding area; inputting the interpolated and expanded temperature matrix into a decision tree model, outputting predicted position information of the smoke source, and calculating spatial position coordinates; converting the point cloud data of the welding area into a grey-scale map, fusing the grey-scale map and the interpolated and expanded temperature matrix data to obtain a temperature space distribution matrix, inputting the temperature space distribution matrix into a first convolutional neural network, and outputting obstacle avoidance point cloud data; calculating parameters of the obstacle avoidance area, inputting the parameters, the spatial position coordinates and the initial joint rotation angle of the actuator into a second convolutional neural network, and outputting a target joint rotation angle; therefore, the welding smoke in the welding area can be automatically sucked by the actuator, the sucking efficiency is improved, the smoke concentration in the welding working area is reduced, the conflict condition of the smoke sucking disc working area is avoided, and the interference to the welding work and the sucking leakage rate are reduced.