Stock trading and prediction using neural networks

• Main Author: Guo, Meng.
• Other Authors: Wang Lipo
• Format: Final Year Project
• Language: English
• Published: 2010
• Subjects: DRNTU::Engineering
• Online Access: http://hdl.handle.net/10356/40600
• Institution: Nanyang Technological University

This paper investigates the method of predicting stock price trends using rule-based neural network which was initially proposed by Seng-cho Timothy Chou, Chau-chen Yang, Chi-huang Chen and Feipei Lai in their paper “A Rule-based Neural Stock Trading Decision Support System” [27]. Artificial neural network (ANN) has one input layer, one hidden layer and one output layer for supervised learning and prediction. The neurogenetic model is trained by input features, which are derived from a number of technical indicators being used by financial experts. After this, a new set of test data will be put into the model for prediction. The genetic algorithm (GA) optimizes the NN’s weights in the mean time. The output from the neural network will be used to make trading decision based on the trading rule and threshold value determined. By testing the proposed method with 18 companies in NYSE and NASDAQ for 10 years from 1999 to 2009, an encouraging result has been showed.