Predicting stock dividend using neural network and decision tree and comparing them with voting technique

Abstract

Data mining is one of the developing sciences and it is very suitable for analyzing database. Data mining is used in many sciences, such as business intelligence, shopping basket analysis, and medicine. The main algorithms of data mining are 4 categories that 2 main categories are feature ranking and classification algorithms. In this study, we propose a method for predicting the dividend of market price using data mining technique. A new method has been provided for classifying the data, and then the accuracy of each method has been achieved by implementing the above approach on a database with 371 companies in different industries and obtaining the precision of each model according to the inputs. We used stock data to apply this research, and based on the approach, we predicted the rate of change in the companies' dividends in 2015 according to the data of companies. The results indicated the high accuracy and high speed of the proposed approach.