5results about How to "Resolving heterogeneity" patented technology

The present application relates to the technical field of artificial intelligence and federated machine learning, and particularly relates to an asynchronous federated learning aggregation method and system based on clustering cache. The present application clusters clients by intermediate features of the clients. In global aggregation, intra-cluster aggregation is firstly performed, and then global aggregation is performed based on client clusters. The client clusters are dynamically updated. In intra-cluster aggregation, an active set and a slow set are introduced to realize an asynchronous participation mechanism of intra-cluster aggregation. Weighted aggregation of the active set and the slow set solves the problem of system heterogeneity. The present application solves the defect of insufficient timeliness of existing federated learning methods in a data heterogeneous scene, and guarantees the timeliness and model training accuracy of federated learning in the data heterogeneous scene.

The invention discloses a construction method and application of a safety evaluation risk prediction model after child medication marketing, and the method comprises the steps: constructing a three-level index system, and determining the weight of each level of index through an analytic hierarchy process; based on the three-level index system, determining a risk feature set, constructing a three-level model architecture, and performing model training by adopting a machine learning algorithm; collecting multi-source real world data, and establishing a standardized data pool; the performance of the model is improved through hyper-parameter optimization and integrated learning, the performance of the model is verified and evaluated by adopting internal cross validation and an external independent data set, and the transparency of the model is enhanced through an interpretability technology; and establishing a dynamic updating mechanism, deploying the model as a cloud API service, and realizing risk prediction and early warning functions. According to the invention, multi-source real world data can be effectively integrated, the children medication safety risk can be accurately identified, the method is used for safety evaluation after children medication is listed, and the method is of great significance to supplement clinical test insufficiency before children medication is listed, reduce medication risk and guarantee children medication safety.

The application discloses a kind of based on retrieval enhancement cross-modal condition diffusion model emotion recognition method and system, the method receives text, speech, image multimodal data, detects modality integrity and extracts existing modal feature;If there is missing mode, based on existing modal feature in external sample library retrieval similar sample, weighted fusion generates retrieval enhancement representation;Cross-modal condition diffusion model is constructed, and missing modal feature is recovered by combining forward diffusion into information, reverse diffusion and iteration unified cross-modal attention mechanism;Splice complete modal feature, pass through Transform fusion network and joint loss function optimization, output emotion category and intensity.The application breaks through the limitation that existing generation model only relies on local information, fully utilizes the combination advantage of external retrieval and diffusion generation, can maintain stable performance under a variety of missing patterns, and has wide application prospect.

The application relates to the technical field of marine environment monitoring, in particular to a marine ecosystem monitoring system based on multi-source data fusion, which comprises a multi-source data acquisition module, a parameter extraction module, an ecological driving analysis module and an ecosystem anomaly monitoring module; wherein: the multi-source data acquisition module is used for acquiring multidimensional data; the parameter extraction module generates unified grid data through space-time alignment processing, and extracts dissolved inorganic phosphorus concentration time series data and diatom community spectral feature data sets; the ecological driving analysis module calculates the fluctuation rate and generates diatom community aggregation degree; the ecosystem anomaly monitoring module is used for identifying the abnormal state of the ecosystem and outputting corresponding early warning signals. Through multi-source data fusion and dynamic trend analysis, the application realizes accurate identification and timely early warning of marine ecological abnormal events, and improves the timeliness and discrimination accuracy of ecological monitoring.

The application discloses a method for intelligent identification of submerged oil pollution based on multi-source heterogeneous data fusion, comprising collecting multi-source monitoring data and meteorological data of a submerged oil area, and preprocessing the multi-source monitoring data and the meteorological data; adopting a stereoscopic perception system to perform space-time registration optimization on the multi-source monitoring data and the meteorological data to obtain spatial alignment data, and performing deep multi-modal fusion on the spatial alignment data to obtain fusion data; adopting a graph neural network to simulate oil cluster migration along a flow based on the fusion data to obtain a simulated oil cluster migration path, and performing oil cluster migration game analysis based on an actual oil cluster migration path and the simulated oil cluster migration path to obtain an oil cluster movement migration rule; and constructing an intelligent identification model for submerged oil pollution based on physical information embedding based on the oil cluster movement migration rule, inputting to-be-identified data into the intelligent identification model for submerged oil pollution, and outputting an identification result.