Further study of adaptive supervised learning decision (ASLD) network in stock market

This paper further studies the recently proposed Adaptive Supervised Learning Decision (ASLD) network for trading and portfolio management [5] in two sets of stock market data. One is the Hang Seng index in Hong Kong markets. The other is a portfolio of indexes from six major markets in the world. Being different from the study on foreign exchange rates [5], we find that in generating the trading signals for training the neural network used in the ASLD system, the issue of volatility should be considered important in handling stock market data. Several heuristic strategies are investigated for taking volatility into the consideration in training the ASLD trading and portfolio system. Empirical results are given to show how well these strategies work.