Title :
Moving average crossovers for short-term equity investment with L-GEM based RBFNN
Author :
Cai, Gao-yang ; Ng, Wing W Y ; Chan, Patrick P K ; Firth, Michael ; Yeung, Daniel S.
Author_Institution :
Machine Learning & Cybern. Res. Center, South China Univ. of Technol., Guangzhou, China
Abstract :
The Shenzhen Stock Exchange (SZSE) market is young and energetic. Evidence exists that the returns from emerging markets like the SSE are influenced by a different set of factors than those of developed markets. The Moving Average (MA) crossover technique is one of the popular technical analysis tools used by investors in financial markets. However, not all MA crossovers give accurate predictions of uptrends in stock prices. This motivates us to investigate the use of MA crossovers in short-term investment with Radial Basis Function Neural Network (RBFNN) trained via a minimization of the Localized Generalization Error (L-GEM). Experiments show that the proposed method can yield statistically significant profits when compared with a random investment strategy.
Keywords :
autoregressive moving average processes; investment; radial basis function networks; stock markets; L-GEM based RBFNN; Shenzhen stock exchange market; financial markets; localized generalization error minimization; moving average crossover technique; radial basis function neural network; short-term equity investment strategy; technical analysis tools; Artificial neural networks; Finance; Investments; Machine learning; Neurons; Stock markets; Training; Equity Market; L-GEM; Moving Average Crossover; RBFNN;
Conference_Titel :
Machine Learning and Cybernetics (ICMLC), 2010 International Conference on
Conference_Location :
Qingdao
Print_ISBN :
978-1-4244-6526-2
DOI :
10.1109/ICMLC.2010.5580985
