A novel evolutionary decision support system (EDSS) that iseffective for making profitable trading decisions in the stock market ispresented. A genetic algorithm is incorporated with a sliding windowscheme to effectively estimate the most profitable trading decision,namely buy, hold or sell. Because of its data-driven nature and thegenetic change to the positive course, EDSS bypasses the complicatedsteps of network establishment and subsequent training, and it candirectly deal with the problem of structural instability that plaguestraditional rule-based systems. Empirical tests are conducted on theweighted price index of the Taiwan stock market (TSEWSI). Compared tothe buy-and-hold strategy, the EDSS can achieve a significantimprovement in profit gains
展开▼
