18 results about "Hypothesis verification" patented technology

The invention relates to the field of medical information processing, in particular to a deep learning-based medical multi-source heterogeneous data fusion and knowledge discovery system, which comprises a heterogeneous data management and mapping module, a medical data unified representation module, a medical knowledge mining engine, a knowledge-driven decision support module and a hypothesis verification and visual exploration module, the system utilizes the Riemannian manifold theory to construct a multi-modal medical data unified characterization framework, realizes multi-scale medical knowledge characterization in combination with a curvature perception mechanism, excavates multi-entity association through multi-view learning and a graph neural network, calculates the consistency among diseases, symptoms, signs and genes based on a knowledge graph, provides accurate diagnosis and treatment schemes, and realizes the purpose of accurate diagnosis and treatment. According to the method, the technical problems that medical multi-source heterogeneous data are difficult to fuse, complex association is difficult to mine and knowledge representation is not accurate enough are effectively solved, and the data utilization efficiency, representation quality and diagnosis accuracy are remarkably improved.

The invention discloses a vehicle nameplate identification method, system and device and a storage medium, and the method comprises the steps: carrying out the multi-scale grid partitioning of an input image, extracting features, building a regional characteristic map to discriminate the imaging attribute category, and carrying out the differential image enhancement and fusion to generate a preprocessing image; constructing a multi-hypothesis verification model based on the geometric attribute and the layout relationship of the vehicle identification code to perform rotation correction; constructing a perspective transformation model by utilizing text line geometric constraint, and correcting spatial deformation through adaptive block perspective transformation and affine transformation; structured information is output by adopting a universal recognition and special engine collaborative decision-making mechanism; and finally carrying out image feature generation, repeated verification, semantic consistency verification and block chain evidence storage. According to the method, the problems of poor vehicle nameplate image quality, difficulty in geometric deformation correction and low identification accuracy in a complex imaging environment are effectively solved, and end-to-end high-precision identification from a low-quality image to credible structured information is realized.

The application discloses a multi-knowledge base fusion deduction method based on collaborative decision-making, comprising: performing semantic analysis on a deduction request input by a user to extract key semantic features; activating a target knowledge base in a multi-knowledge base cluster based on the features and retrieving preliminary knowledge fragments; performing semantic alignment and conflict resolution on the preliminary knowledge fragments to generate a fusion knowledge graph; performing cross-library joint reasoning on the fusion knowledge graph, and performing a multi-step deduction cycle with the preliminary reasoning conclusion as a starting point, wherein a hypothesis generation and verification mechanism is introduced; and finally performing confidence evaluation, sorting and synthesis on each conclusion in the final deduction conclusion chain to output an optimal deduction result. The application realizes a leap from "information retrieval" to "knowledge deduction", has advantages such as cross-library joint reasoning, multi-step deduction cycle, hypothesis verification and interpretability, and improves reasoning capability, dynamic adaptability and decision quality in a complex decision-making scenario.

The invention discloses a fault diagnosis and safety early warning system for an automatic driving carrying robot, and the system comprises the steps: collecting panoramic operation tense data, carrying out the threshold judgment, cross validation, performance and trend analysis, and task execution result verification, and outputting a preliminary diagnosis result; performing time anchor point diffusion analysis, locking a core event range, extracting related synchronization fragments, analyzing an association relationship, calculating association strength of different data exceptions and core faults, and locking an exceptional operation data range; constructing an abnormal event graph structure, identifying a root cause node and a key propagation path, generating a macroscopic abnormal event graph, performing hypothesis verification, and outputting an abnormal operation data chain; performing scene reproduction through digital twinning to position a root cause, and outputting a fault analysis result; and performing hierarchical security early warning and generating a fault file, and storing the fault file in a block chain. The method aims at solving the problems that complex faults of an automatic driving carrying robot are difficult to position quickly, and the root cause diagnosis efficiency is low.

This invention relates to the field of artificial intelligence technology and discloses a hypothesis-verification-based method, apparatus, and device for invoking tools by an intelligent agent. The method includes: acquiring a user task; acquiring a preliminary candidate toolset corresponding to the user task based on a preset tool capability graph; each node in the tool capability graph represents a tool; the preliminary candidate toolset is a set of tools required by the user task; each node in the tool capability graph has its corresponding attribute parameters; generating usage hypotheses corresponding to the tools in the preliminary candidate toolset based on the attribute parameters; acquiring expected target hypotheses based on the usage hypotheses; and invoking the tools corresponding to the expected target hypotheses to generate a completion plan corresponding to the user task. This allows for the simulation of the functions corresponding to each task in the preliminary candidate toolset, thereby selecting expected target hypotheses that meet the requirements of the user task, improving the accuracy of the selected tools, and thus increasing the success rate of the generated completion plan.

This invention discloses a high-precision detection method for concave and convex feature points based on multi-segment straight line edge analysis, comprising the following steps: S10, image acquisition; S20, preprocessing and edge detection; S30, straight line segment extraction and filtering; S40, line segment relationship analysis and concave / convex hypothesis generation; S50, hypothesis verification and precise positioning; S60, outputting the coordinates of concave and convex points. This invention overcomes the dependence on continuous and complete contours by directly detecting and filtering multi-segment straight line edges related to potential concave and convex features from the image. Based on this, it robustly and accurately locates the coordinates of convex and concave points by analyzing the spatial topological relationships and geometric constraints between these straight line segments.

An object of the system according to the embodiment is to repeat hypothesis verification at high speed and frequently and to quickly announce a result useful for the industry.SOLUTION: A system includes a data analysis part, a document review part, an experiment design part, a report creation part, and a communication support part. The data analysis unit performs data analysis. The document review unit performs a document review. The experiment design unit proposes an experiment design. The report creator supports creation of a report. The communication support unit promotes communication and cooperation.SELECTED DRAWING: Figure 1

The invention discloses a track defect image segmentation method based on visual knowledge base retrieval and verification, and relates to the field of image segmentation, and the method comprises the steps: based on a query image and a text instruction describing defects in the query image, retrieving at least one track defect image sample related to the query image in the visual knowledge base, taking as visual evidence; carrying out concept matching verification and position hypothesis verification on the visual evidence; the concept matching verification is used for generating a first Boolean judgment result; the position hypothesis test is used for generating a second Boolean judgment result; generating space prompt information according to the first Boolean judgment result and the second Boolean judgment result; and based on the space prompt information, performing defect segmentation on the query image by using a segmentation model. According to the invention, the accuracy of track defect image segmentation is improved.

The invention provides a construction method of an intelligent container collection gravity model of a storage yard. The construction method comprises the following steps: establishing a UCT-MCTS framework; carrying out symbolic modeling on the basis of the UCT-MCTS framework; target driving design is carried out based on the UCT-MCTS framework, and a single-outlet container position selection total target function related to maximization of container attraction, minimization of rejection and maximization of site tool attraction is constructed; hypothesis verification is carried out based on the UCT-MCTS framework, and application conditions of a corresponding model are determined; and carrying out constraint design based on the UCT-MCTS framework. The invention further provides a scheduling method, location selection optimization is carried out on outlet container stockpiling of a container wharf through a storage yard intelligent container receiving gravitational model, and the storage yard intelligent container receiving gravitational model is the UCT-MCTS framework constructed through the construction method. The invention further provides a computer readable storage medium for executing the method. According to the invention, by constructing the multi-dimensional gravity model and the optimization algorithm, accurate and efficient container position selection decision is realized, and the operation efficiency of a container port is improved.

The invention particularly relates to a diagnosis and treatment data management system based on cell therapy and negative oxygen ion conditioning, and relates to the technical field of medical data processing, comprising the following steps: generating a potential root cause hypothesis list sorted according to probabilities, and aiming at high probability hypotheses in the potential root cause hypothesis list, selecting a potential root cause hypothesis list; and automatically calling, associating and analyzing corresponding original environment data logs and cell preparation process data before and after the abnormal time point so as to carry out hypothesis verification and final root cause positioning. According to the method, a dynamic sampling, high-precision time synchronization and data quality monitoring mechanism is constructed, and a high-quality data basis is provided for anomaly analysis; the real-time anomaly detection and intelligent root cause analysis module adopts a multi-algorithm parallel and voting grading early warning strategy, and adapts corresponding algorithms for different data types; and during the second-level early warning, through Bayesian reasoning accelerated by the GPU, only traversing upstream nodes of the abnormal indexes, completing posterior probability calculation, and combining the posterior probability and the influence range to solve difficulty comprehensive sorting root causes.

The application provides a construction method of a yard intelligent container receiving attraction model, comprising: establishing a UCT-MCTS framework; performing symbolic modeling based on the UCT-MCTS framework; performing target driving design based on the UCT-MCTS framework, and constructing a single-outlet container positioning total target function related to maximum container attraction, minimum repulsion and maximum yard tool attraction; performing hypothesis verification based on the UCT-MCTS framework, and determining the applicable conditions of the corresponding model; and performing constraint design based on the UCT-MCTS framework. The application also provides a scheduling method, which optimizes the positioning of outlet container storage of a container terminal through a yard intelligent container receiving attraction model, wherein the yard intelligent container receiving attraction model is a UCT-MCTS framework constructed by the construction method described above. The application also provides a computer readable storage medium for executing the above method. Through the construction of a multi-dimensional attraction model and an optimization algorithm, the application realizes accurate and efficient container positioning decision-making, and improves the operation efficiency of a container port.

The invention discloses a knowledge fusion method and device, a machine readable storage medium and electronic equipment. The knowledge fusion method comprises the following steps: determining an unmatched node set of a term standardization transition network based on a mapping rule base; determining at least one candidate interpretation hypothesis for each first node in the determined mismatched node set; verifying the at least one candidate interpretation hypothesis, and adjusting the mapping rule base according to a verification result; repeatedly executing the steps of determining an unmatched node set, determining at least one candidate interpretation hypothesis, verifying the at least one candidate interpretation hypothesis and adjusting the mapping rule base until a preset condition is met; and outputting the final mapping rule base. Therefore, the translation dilemma between different knowledge fields is relieved so as to promote the combined research between different knowledge fields.

The invention relates to the technical field of medicine, in particular to a large health medical model construction method, which comprises the following steps of: through six-step closed loop of'ancient book hypothesis-CRISPR verification-digital twinning-cross-species mapping-heterologous fusion-reinforcement learning ', converting leaflet-level ancient book information into quantifiable biological parameters with confidence greater than or equal to 92% for the first time; the cross-species data fusion error decrease is greater than or equal to 32%, and the model AUC increase is greater than or equal to 0.28; the organ-level 4D physiological matrix can directly drive any AI frame and reinforcement learning sandbox to output an executable and explainable personalized space health scheme within a minute level, so that bone loss is greater than or equal to 40%, a radiation damage index is greater than or equal to 30%, intervention cost is greater than or equal to 25%, the scheme can be seamlessly migrated to the ground for precise medical treatment, and an'ancient book-gene-clinical 'integrated closed loop is constructed. Therefore, the problem that no systematic method for combining ancient biological intelligence with the modern medical technology exists at present is solved.

The invention relates to a fault diagnosis method and system, a readable medium and electronic equipment, and relates to the technical field of system operation and maintaining.When alarm information is triggered, the scheme can plan a first data collection task to obtain preliminary evidence of the alarm information, and a first prompt word is constructed on the basis of the preliminary evidence; calling the first large model to carry out hypothesis reasoning to obtain a fault hypothesis; and further, planning a second data acquisition task based on the fault hypothesis to obtain key evidence, constructing a second prompt word based on the preliminary evidence, the fault hypothesis and the key evidence, calling a second large model to carry out verification reasoning to obtain a fault root cause, and comprehensively obtaining a fault diagnosis result. According to the scheme, full-amount data does not need to be deeply analyzed, and the requirement for computing resources is reduced; in the assumption-verification process, collection and analysis can be guided through an intermediate conclusion, and noise interference is reduced; causal association is fully mined based on a large model, the reliability is improved, the method also has good adaptability to complex dynamic scenes, and the generalization ability is improved.

The invention relates to a deep neural network internal neuron semantic interpretation and verification method, and belongs to the technical field of artificial intelligence and computer vision. The method comprises the following steps: giving a to-be-explained target deep neural network and a probe data set; performing activation distribution analysis on neurons in the network, and screening out target neurons with high distinction degree and corresponding strong activation samples based on statistical characteristics of activation values; analyzing the strong activation sample by using a clustering algorithm and a pre-trained vision-language model, and generating a semantic concept hypothesis of a neuron function; generating an image model by using a text, and generating a verification image set according to the generated semantic concept hypothesis; and inputting the verification image set into the target deep neural network, and verifying the correctness of the semantic concept hypothesis by calculating the activation rate of neurons on the verification image set. According to the method, through a closed-loop framework of screening-hypothesis-verification, the explanatability accuracy and credibility of the deep learning model are improved.

The application discloses a forestry intelligent pest control system based on multi-sensor cooperation and relates to the technical field of intelligent pest control. In order to solve the problems of low diagnosis accuracy and poor cooperation efficiency in the prior art, the application proposes a "hypothesis-verification" cooperation strategy. The system first generates a preliminary diagnosis hypothesis by using macro-sensor data and then adaptively adjusts and controls micro-sensor for targeted verification according to the preliminary diagnosis hypothesis, so as to realize high-confidence diagnosis. The system further establishes a pest spatial-temporal diffusion risk map to realize trend prediction and formulates a graded and zoned precision control strategy. Through closed-loop self-optimization learning of the prevention and control effect evaluation and the model, the application realizes intelligent control from passive response to active prediction, and significantly improves the diagnosis accuracy and resource utilization efficiency.

The invention provides a rare disease scene interactive differential diagnosis method and device, and belongs to the technical field of intelligent medical treatment and auxiliary diagnosis, and the method comprises the steps: obtaining an unstructured symptom text of a patient, and generating a preliminary hypothesis phenotype and candidate diseases through a generative large language model; constructing a multi-source heterogeneous graph network, performing knowledge constraint reasoning by using a graph neural network, and outputting disease probability distribution and phenotype associated information; constructing a multi-factor causal matrix based on the causal association of the anti-fact intervention quantification phenotype and the disease; and optimizing an interaction strategy through reinforcement learning, dynamically selecting a key phenotype to generate an inquiry, and iteratively updating the phenotype group of the patient until the patient is diagnosed. According to the method, hypothesis-verification-optimization closed-loop diagnosis is realized, the diagnosis accuracy and efficiency are improved, the reasoning process is traceable and explainable, and the method is suitable for a rare disease clinical auxiliary diagnosis scene.

A computing system including a processing circuit in communication with a camera having a field of view. The processing circuit is configured to perform operations related to detecting, identifying, and retrieving objects disposed amongst a plurality of objects. The processing circuit may be configured to perform operations related to object recognition template generation, feature generation, hypothesis generation, hypothesis refinement, and hypothesis validation.