استدلال روش جدید داده کاوی ترکیبی با استفاده از یک مورد مبتنی بر رگرسیون : کاربرد برای پیش بینی مالی

This paper proposes a regression case based reasoning (RCBR) which applies different weights to independent variables before finding similar cases. The traditional CBR model has focused on finding similar cases from a case base without considering the importance of independent variables. Thus, when extracting similar cases the traditional CBR has put same weights on each independent variable. The proposed regression CBR (RCBR) finds a relative importance of independent variables from the relationship between independent variables and a dependent variable using a regression analysis and puts relative weights using regression coefficients on independent variables. Then, it selects nearest neighbor or similar cases using weighted independent variables through the traditional CBR machine and updates weights dynamically for the next target case and again performs the traditional CBR machine.

This paper proposes a new hybrid learning technique, a regression CBR (RCBR) which applies different weights to independent variables before finding similar cases. The traditional CBR model has focused on finding similar cases from a case base without considering the importance of independent variables. Thus, when extracting similar cases the traditional CBR has put same weights on each independent variable. The proposed regression CBR (RCBR) finds a relative importance of independent variables from the relationship between independent variables and a dependent variable using a regression analysis and applies relative weights using regression coefficients to independent variables. Then, it selects nearest neighbor or similar cases using weighted independent variables through the traditional CBR machine and updates weights dynamically for the next target case and again performs the traditional CBR machine. These concepts are investigated against the backdrop of a practical application involving the prediction of a stock market index. The rest of this paper is organized into five sections. Section 2 reviews case based reasoning (CBR) as a knowledge discovery technique. Section 3 introduces a regression CBR. Section 4 presents the case study and reports the results. The case study intends to investigate the effect of a regression CBR on the predictive performance in forecasting a stock market index. Finally, the concluding remarks are presented in Section 6.

In recent years, data mining techniques such as neural networks, CBR have been applied extensively to the task of predicting financial variables. This paper proposed a new case based reasoning technique which concepts are investigated against the backdrop of a practical application involving the prediction of a stock market index. The regression CBR was significantly better than random walk and was better than standard CBR models in the hit rate measure as well as was seen to surpass other models in the MAPE. In fact, the literature contains numerous varieties of learning techniques such as neural networks, induction and genetic algorithms. Thus, the systematic evaluation of a larger collection of learning techniques represents a rich area for future investigation. Also a promising direction for the future is a trading strategy through integrating a regression CBR and other learning techniques for financial market.