Predicting stock trends through technical analysis and nearest neighbor classification
SMC'09 Proceedings of the 2009 IEEE international conference on Systems, Man and Cybernetics
Expert Systems with Applications: An International Journal
Hi-index | 0.00 |
Making profit based on historical data is one of the most important problems in the study of financial markets. This paper presents a method which selects the top twenty, seventeen and fifteen stocks by choosing the top five stocks from each equity fund. The equity funds are selected by performance evaluation and the PSO algorithm is applied to allocate the investment capitals of stocks_ portfolios. In this paper, we compare the accumulative return rates of constructed portfolios with the Taiwan Weighted Stock Index and of the best Equity Fund in each month. The results show that the PSO algorithm is effective in making better returns of portfolios.