A Novel Regression Method Research Based on Covering Algorithm

The Covering algorithm is proposed by Professor ZhangLing and ZhangBo in the 20th century, which simulates the structure of human learning, building a Constructive Neural Network Learning Model. Covering algorithm has been widely used to solve massive data classification problem, because its performance. The covering classification algorithm has fast learning, high recognition rate, massive data processing capacity and clear physical meaning advantages which has been used into the field of Weather forecasts, financial analysis and so on. So we try to inherit and develop the advantage of covering classification algorithm and expand its applications to the regression prediction.