Hybrid Intelligent Decision Support Systems for Selection of Alternatives in Stock Trading

The dissertation presents Hybrid Intelligent Decision Support Systems (DSS) for the selection of alternatives (companies, stocks, and company groups) in stock trading under uncertainty. This study proposes a framework including three models using Hybrid Intelligent DSS. The framework aims to optimize trading decisions in the selection of appropriate alternatives and reduce risky decisions. This framework is used to quantify qualitative attributes and normalize quantitative attributes of alternatives, together with expert preferences and sensibilities under uncertainty for the selection of alternatives and reducing risks in stock trading. To validate the performance of this framework, the proposed models in the framework have been tested and performed objectively by multiple experts in real-world stock trading through experiments in case studies on the HOSE, HNX and NYSE and NASDAQ stock markets. In this framework, the first model called a Hybrid SOM-AHP model is a SelfOrganizing Map (SOM) integrated with Analytic Hierarchy Process (AHP). This model aims to select short-list investment alternatives in rankings for stock trading. Experimental results of this model showed an average rate from 68% to 70% in stock selection for successful investment. The second model called Hybrid Kansei-SOM (HKS) model is integrated by SOM with Kansei evaluation for quantifying expert sensibilities in trading decisions. The experimental results showed that HKS model obtained successfully stock selection rate from 81% to 85% in investments. The third model called Hybrid Kansei-SOM Risk (HKSR) model aims to reduce risky decisions and alternative risks. Compared to HKS model, HKSR model was reduced risky stocks in investments from 3% to 5% better than that of HKS model. Compared with