A data classification method and device based on a discrete dynamic mechanism

Abstract

The invention discloses a data classification method and device based on a discrete dynamic mechanism, and the device is based on the data classification method based on the discrete dynamic mechanism, and the method comprises the steps: receiving a data sample set, and carrying out the data preprocessing; constructing a dynamic neural network model, wherein each layer of the dynamic neural network model comprises a plurality of dynamic modules established by neurons, the dynamic module in each layer is connected with a certain dynamic module in the next layer, and a full connection network isformed between features extracted by the dynamic module in the last layer and data types; Using the standardized data sample set to train a dynamic neural network model to obtain a data classification model; And receiving to-be-classified data, and carrying out data classification according to the data classification model.

Description

technical field [0001] The disclosure belongs to the technical field of data processing, and relates to a data classification method and device based on a discrete dynamic mechanism. Background technique [0002] The statements in this section merely provide background information related to the present disclosure and do not necessarily constitute prior art. [0003] In today's era, data has become an important resource. It is very meaningful to analyze and process the existing data, find out the rules and classify them. [0004] Due to the limitations of hardware performance, the effect of deep learning models is not ideal. In recent years, the emergence of graphics processing units (GPUs) has set off a new wave of deep learning and promoted the further development of artificial intelligence. In almost all deep models, the convolutional neural network module (CNN) plays a key role in the analysis and processing of big data. At present, almost all deep learning models will...

AI Technical Summary

Problems solved by technology

[0006] Aiming at the deficiencies in the prior art and solving the problems of unsatisfactory data classification performance, slow speed and low accuracy caused by the insufficient use of data correlation in the convolutional neural network and too many model parameters in data classification, this disclosure One or more embodiments of the present invention provide a data classification method and device based on a discrete dynamic mechanism, making full use of the dependencies between information, so as to better mine the internal information of the data for data classification, suitable for text data, Classification of data such as image data and audio

Images