Reversal Pattern Discovery in Financial Time Series Based on Fuzzy Candlestick Lines

This paper provides a new technical analysis method for finding reversal points of stock price. The proposed method applies the Fuzzy logic theory to the Japanese candlestick theory by converting the open, close, high and low prices into fuzzy candlestick chart. Using the model, we can define and interpret the “symptom sequence” before the reversal point, and then identify the reversal patterns of the candlestick chart. Transaction data of non- ST Shares in SSE A Share, SSE B Share, Shenzhen A share and Shenzhen B share are analyzed to validate the proposed model, respectively. Results demonstrate that reversal patterns are convincingly identified by the model, and basically the model could be applied to establish a stock price reversal early warning system in the financial market.