Reinforcement learning representation method for continuous-categorical mixed data classification
The rapid development of artificial intelligence technology has detonated the era of big data,resulting in various types of data,data drives the development of machine learning,and the performance of machine learning also depends on the representation of such data by the representation model,while the traditional representation algorithm can not make the best distinction of such data.In order to solve the above problems,this paper introduces reinforcement learning on the basis of the representation algorithm,takes the clustering evaluation index as the reward,obtains the best distinguishing representation data,and applies it to the classification task.Experimental results show that the proposed reinforcement learning representation method can achieve better results in machine learning classification tasks than traditional representation methods.
万方数据
维普
