Document Type



While there are a number of finance methods (fundamental analysis, technical analysis, contrarians’ theory, risks management, etc) used in stock markets to help make investment decisions, they have different strengths and weakness. It is observed that these different finance methods are not being integrated by existing technologies in a systematic way, thus their performance for identifying investment opportunities is limited. In this research, I propose a systematic method (i.e. an IDSS system) to take advantage of and to optimize existing and newly proposed methods in order to obtain better investment performance, through identification and classification of Positive, Neutral and Negative investment opportunity ranges and related risks. This IDSS system will be mainly based on Turning Point Model and Optimized AutoSplit method, which help find hidden investment opportunities and risk variables, particularly a stock’s unique trend. The key methodology is to use Decision Tree theory with finance knowledge. The IDSS system will be built on the top of F-trade platform which has been already developed by UTS Data Mining team and has a RDP structure, with agent-based distributed expert systems. Initial system evaluation shows that the system successfully identified investment opportunity ranges, outperforming the benchmark index and other systems.