Hierarchical Classification Storage of Data Based on Deep Adaptive Clustering Algorithm
In order to improve the power data information management and control ability,a new data hierarchical classification storage method is proposed.This method adopts deep adaptive clustering algorithm,which can not only recognize the features of single data effectively,but also recognize the whole structure of data to a certain extent.By autoencoder DNN model and graph neural network GCN model,the hierarchical classification of complex data can be realized efficiently.The data storage system based on algorithm can convert various kinds of data information of power data information into digital information.The deep learning algorithm is used to analyze the internal connections of data,improve the storage capacity of the system.Experi-mental results show that the classification accuracies of low data volume,complex data volume and high data volume can reach 97.5％,92％and 86％,respectively.The data classification efficiency can reach about 97％.
万方数据
维普
