DEEP REINFORCEMENT LEARNING-BASED DATA CLASSIFICATION METHOD, APPARATUS, DEVICE, AND MEDIUM

摘要

Provided are a deep reinforcement learning-based data classification method, apparatus, device, and medium, relating to the technical field of artificial intelligence, and being specifically applicable to deep learning. The method comprises: formatting acquired survey data to obtain initial data; pre-processing classification features to obtain training data; training result information by means of training data and a feature solution to generate feature values; using the feature values to adjust the parameters of the training data and the parameters of the feature solution, and taking the adjusted training data and feature solution as a target solution; by means of a deep reinforcement learning model, using the target solution to iteratively calculate result information to obtain a clustering decision model; inputting the acquired feature data of a user into the clustering decision model to obtain a classification corresponding to the user. The present application also relates to blockchain technology; the survey data is stored in the blockchain. By means of training the clustering decision model, the accuracy of data classification of diseases is improved.