Maximizing winning trades using a rough set based other-product (RSPOP) fuzzy neural network intelligent stock trading system

Trading systems have been relying more and more on the use of novel computational intelligence techniques in the formulation of trading decisions. A novel RSPOP intelligent stock trading system is proposed in this paper. This trading system is demonstrated empirically to achieve significantly superior returns on live stock data, and is able to filter out erroneous trading signals generated by the moving average trading rule. This ability to filter out erroneous signals is measured by the percentage of winning trades. The trading system is demonstrated empirically to achieve more than 92% of winning trades compared to an average of 70% of winning trades demonstrated by the conventional trading system based on the moving average trading rule.