Advanced Stock Price Prediction Using Facebook Prophet-Arima Ensemble Model

Accurate stock price forecasting has become crucial at a time of economic uncertainty and significant market volatility. It is frequently difficult for conventional financial models like the Exponential Moving Average, Relative Strength Index (RSI), Moving Average Convergence Divergence (MACD), and Dow Theory to account for the non-linear character of stock market cycles. Contemporary machine learning techniques, such as the ARIMA and Prophet models. This paper creates a predictive framework that combines Prophet's capacity to manage seasonality, outliers, and trend fluctuations with ARIMA's efficacy in short-term forecasting. We put these models into practice via a Streamlit-based application that retrieves current stock data from Yahoo Finance, analyzes it, and forecasts future prices. While Prophet is used for long-term trend forecasting, the ARIMA model is used to anticipate short-term price fluctuations. In order to increase forecast accuracy, we also provide an ensemble strategy that combines predictions from both models. Performance is evaluated using model assessment measures such as the mean absolute error (MAE), the mean squared error (MSE), the root mean squared error (RMSE) and the mean absolute percentage error (MAPE). We examine how well these models forecast price patterns for well-known equities like Google (GOOG), Apple (AAPL), Microsoft (MSFT), Tata Consultancy Services Limited (TCS.NS) and Bharat Electronics Limited (BEL.NS) using publicly available stock data. The study's conclusions bridge the gap between conventional financial indicators and AI -driven forecasting methods, offering investors and quantitative analysts insightful information.