This project delves into the field of sales forecasting using time series analysis, focusing on evaluating and comparing the forecasting accuracy of four models: ARIMA, SARIMA, GRU, and Prophet. The primary objective is to determine the model most capable of delivering precise sales predictions, which is critical for optimizing sales and inventory management strategies. Performance evaluation is conducted using key metrics such as Mean Squared Error (MSE), Root Mean Squared Error (RMSE), and Mean Absolute Error (MAE). The project involves a comprehensive workflow that includes data preparation, model training, and rigorous performance assessment. By identifying the most accurate forecasting model, this research aims to assist businesses in crafting flexible, data-driven strategies that can adapt to fluctuating market conditions. Ultimately, the insights generated from this study empower companies to enhance their decision-making processes, boosting operational efficiency and strategic resilience.

Keywords: Sales Forecasting, Time Series, ARIMA, SARIMA, GRU, Prophet, Mean Squared Error (MSE), Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), Performance Metrics, Data-Driven Insights, Inventory Management, Forecasting Accuracy.