New Linkage Learning Technique in Genetic Algorithm for Stock Selection

R. Lakshmi, K.Vivekanandan, S.Amilan

Abstract


In the present paper the researchers develop a new Linkage Learning method draws its concept from machine learning approach and incorporated it with the existing Genetic Algorithms, which would identify a stock with better performance among the given number of stocks. This new Linkage Learning Genetic Algorithm uses historical\fundamental financial indicators and price information of stock trading. Using that information the process would find a stock that can be graded for its performance based on the expected returns. In several studies on stock market the problem of identifying the good stock have been solved by Hidden Markov Model, Artificial Neural Network, Simulated Annealing and Simple Genetic Algorithm. However the linkage Learning Genetic Algorithm has not been applied in the area of stock market for the problem dealt in the present study. Therefore an attempt has been made to use this new Linkage Learning process to identify high quality stocks with investment value using financial indicators. Experimental results with real data from the Indian Stock Market reveal that the Linkage Learning Genetic Algorithm for stock selection leads to better results.

 


Keywords: Linkage Learning, Association Rule Learning, Genetic Algorithms, Stock performance forecast.


Full Text:

PDF


DOI: https://doi.org/10.26483/ijarcs.v5i3.2054

Refbacks

  • There are currently no refbacks.




Copyright (c) 2016 International Journal of Advanced Research in Computer Science