9 results about "Data approximation" patented technology

One example method includes receiving a Quadratic Unconstrained Binary Optimization (QUBO) problem that comprises a matrix that includes various cells having data. It is then determined that one or more of cells is missing data or has corrupted data. A machine learning (ML) model performs a denoising process that removes random noise from the one or more cells having the missing data or corrupted data. This results in data being imputed to the one or more cells having the missing data or the corrupted data. The imputed data approximates the missing data or approximates an expected value of the corrupted data before the corrupted data was corrupted.

The invention discloses a fan blade defect detection system and method based on artificial intelligence, and relates to the technical field of fan blade defect detection.The method comprises the steps that the detection reference value of historical fan blades to the fan blades is evaluated; evaluating a data consistency state of imaging parameters among target historical fan blades in historical blade imaging data, analyzing a data approximation condition of marked imaging parameters among the target historical fan blades, and evaluating an influence condition of the marked imaging parameters on fan blade imaging under different numerical spans; obtaining marked imaging data; according to the marking imaging data and the application imaging data of the fan blade, the fan blade is shot in a domestic power plant to obtain a fan blade image, and defect detection is performed on the fan blade according to the fan blade image to obtain defect detection data, so that the reflecting capability of the image to the real condition of the fan blade is fundamentally improved; and the accuracy of fan blade defect detection is improved.

The application provides an SRAM-based digital domain in-memory computing method and device. First, input data and bit high-low order of the input data are acquired, and SRAM in-memory computing arrays corresponding to the input data respectively are determined according to the bit high-low order of the input data. Different preset approximate adder trees are included in different SRAM in-memory computing arrays, the different preset approximate adder trees include preset approximate full adders with different calculation accuracies, and the structure of the adder tree is a Wallace tree structure. Finally, the input data are subjected to approximate operation through the preset approximate adder tree, and a data approximate operation result is obtained. Through the above method, in combination with the application of the preset approximate full adder and the Wallace tree structure, the preset approximate adder tree with different accuracies can be selected to perform data calculation according to the bit high-low order of the input data while reducing the area and power consumption of the adder tree, and the calculation accuracy is improved.

A learning device (100) includes a defective part shape data storage unit (107) for storing defective part shape data, generated from a new defective product image by a trained model constructed by a learning model generation unit (105), in association with the new defective product image. The image processing device (200) includes: a defective part shape addition unit (207) for extracting defective part shape data, corresponding to a defective product image approximate to input product shape data, from the defective part shape data storage unit (107) and adding the defective part shape data to the product shape data; and a defective product image generation unit (210) for generating a defective product image by optical simulation based on inspection illumination information and the product shape data including the added defective part shape data.

The invention discloses a low-complexity large-model long-sequence memory modeling method. The method comprises the following steps: firstly, constructing a fractional order state space model FOSSM by introducing a Capto fractional order derivative to replace an integer order differential operator; then, based on an asymptotic behavior of a Mittag-Leffler function, establishing a theoretical relationship between a power law attenuation property of an FOSSM memory kernel and a fractional order parameter; deducing the theoretical advantage boundary of the FOSSM in the aspect of power law data approximation compared with the standard SSM; and finally, the non-rational fractional order operator is converted into a rational transfer function through an Oustaloup rational approximation method, and efficient discretization of calculation complexity is realized. According to the method, power law long-range dependence can be modeled naturally, more matched mathematical representation is provided for a sequence with heavy-tailed time correlation, and the problem of path repetition caused by insufficient historical track memory in unmanned equipment navigation and the problem of poor interaction continuity caused by weak context keeping ability in man-machine interaction in a traditional method are solved.

This invention provides a method for estimating the health status of lithium-ion batteries based on Gibbs Variational Inference Multiple Imputation (GVIMI). The method includes: constructing a lithium-ion battery degradation model with missing data; transforming the data imputation problem into a posterior approximation problem based on Gibbs variational inference; optimizing the variational lower bound using a mini-batch data approximation method; obtaining an approximate posterior distribution by maximizing the variational lower bound; and establishing a multiple imputation and health status prediction model. This invention uses a hybrid method combining Gibbs sampling and variational inference as the underlying imputator, introduces a variationally complete conditional model, and uses Gibbs variational inference to approximate the posterior distribution of the model's latent variables, optimizing the imputation model with missing data. This improves the accuracy and robustness of lithium-ion battery health status estimation in scenarios with missing data.

The application discloses a data classification method and device. One embodiment of the method comprises: determining knowledge data under a target coarse classification to which to-be-classified data belongs; determining target historical data similar to the knowledge data; in response to determining that the target historical data includes unclassified data, determining a target fineness of a classification standard to be adopted by a classification standard knowledge base according to an accuracy rate of a classification result of historical to-be-classified data; determining and storing a classification result of the unclassified data by a large language model according to the classification standard knowledge base adopting the classification standard of the target fineness, and the fineness of the classification standard of the classification standard knowledge base is improved in stages; and determining a classification result of the to-be-classified data according to the classification result of the target historical data. The application can obtain a classification result with high precision and high accuracy by using low-precision and easily-obtained data resources such as knowledge data under coarse classification, historical data and a classification standard knowledge base, and the fineness of the classification standard of the classification standard knowledge base is improved in stages, so that the precision and accuracy of the classification result of the to-be-classified data are continuously improved over time.

The invention relates to the technical field of medical data processing, in particular to a medical-field-oriented multi-modal medical data approximate query processing method and system, and the method comprises the steps: detecting the multi-modal type of medical data, and selecting a corresponding query strategy according to the data type and data distribution characteristics; mapping the data to a unified semantic vector space by using a zero sample generation model, and calculating cross-modal similarity to generate a preliminary query result; taking the preliminary query result as input, constructing a multi-target optimization algorithm taking query precision, response time and resource consumption as optimization targets based on an error range, calculation time limitation and calculation overhead set by a user, performing refined screening on the preliminary query result, and dynamically adjusting a query strategy; and receiving feedback information of a user on a query result, and dynamically adjusting model parameters and a query strategy to realize continuous optimization of query performance. According to the invention, through similarity comparison, dynamic threshold optimization and doctor feedback closed loop, high precision and fast response of medical query are realized.