An implementation of knowledge based pattern recognition for financial prediction

Japanese candlestick charting and analysis is one of the most widely used technical analysis techniques and definitely viable and effective for stock and commodity market timing and analysis. The candlestick theory is an empirical model of investment decision. The candle patterns reflect the psychology of market, and the investor can make his investment decision by the identified candle patterns. A knowledge based pattern recognition method for candlestick pattern is proposed in this paper. The investor can define the imprecise and vague candle pattern by this method. The investing expertise can be stored in the knowledge base and the modeled candle pattern can be identified automatically from the large amount of the trading data by the software tool. A prototype system has been established for automating the process of candle pattern recognition.