Development of an investment modeling framework based on the integration of Japanese candlestick patterns and fuzzy logic

Abstract

This research aims to develop an investment modeling framework based on the integration of Japanese candlestick patterns and fuzzy logic. The simultaneous use of these two approaches allows for deeper analysis of historical stock market data and more accurate predictions of future trends. In the proposed model, key indicators such as the Relative Strength Index (RSI) and trend reversal are considered as the main variables, and the data is displayed in a clear and interpretable manner through candlestick patterns. Clustering-based fuzzy rules were designed to explain the characteristics of sequential patterns and quantify the behavior of trends. The use of fuzzy information retrieval mechanisms along with market movement analysis improves the predictive power of the model and increases the accuracy in identifying future trends. The results of the implementation show that combining fuzzy logic with candlestick patterns can be an efficient tool for simultaneously extracting quantitative and qualitative aspects of time series data and providing more stable and reliable forecasts in the stock market. In the proposed model, a new aggregation operator was designed that, while managing the uncertainty and complexity of the data, was able to overcome the limitations of traditional statistical and computational inference systems.