10.55041/IJSREM47795

Accurate stock market forecasting remains a challenging task due to the unpredictable and volatile nature of financial data. Traditional methods like candlestick chart analysis have helped traders interpret market trends visually, but they often lack predictive capability. In this study, we propose a hybrid model that integrates candlestick pattern recognition with the Prophet algorithm— a time-series forecasting tool developed by Facebook. The model is trained on historical OHLC (Open, High, Low, Close) data from the NIFTY 50 index, with an emphasis on capturing both short- term volatility and long-term seasonality. . By detecting key candlestick patterns (e.g., Doji, Engulfing, Hammer) and combining them with Prophet’s seasonality-aware forecasting engine, the system generates actionable buy/sell signals. The results indicate improved prediction accuracy compared to traditional pattern-only methods. This approach bridges the gap between classic technical analysis and modern AI forecasting, offering a more data-driven and adaptable tool for traders and investors operating in the Indian stock market. Key Words: NIFTY 50, Candlestick Patterns, Prophet Algorithm, Time-Series Analysis, Technical Indicators, OHLC Data.