PREDICTION BASED RETAIL MARKET DEMAND FORECASTING

Mr. Srikanth Ganta, Professor, Information Technology, MVGR College of Engineering. Viswanadham Niharika, Information Technology, MVGR College of Engineering, Pothabattula Syamala, Information Technology, MVGR College of Engineering,

Guruvugari Naveen Satya Pydiraju, Information Technology, MVGR College of Engineering, Teeti Nirmal Sai Kumar, Information Technology, MVGR College of Engineering.

***

Abstract – Retail sales forecasting plays a crucial role in helping businesses make data-driven decisions related to inventory management, demand planning, marketing strategies, and financial planning. Accurate forecasting enables retailers to anticipate customer demand, optimize stock levels, minimize losses, and maximize profit margins. This project focuses on developing a predictive analytics system for retail sales forecasting using historical sales data and relevant features. The proposed system utilizes advanced machine learning algorithms, particularly the XGBoost Regressor, to identify complex patterns in sales data and generate accurate future demand predictions. The system is implemented as a web-based application that allows users to upload datasets, visualize historical and predicted trends, and obtain forecast results in both graphical and tabular formats. It also incorporates secure user authentication using Supabase, along with functionality for generating PDF reports and sharing them via email and WhatsApp with password protection. The system improves operational efficiency and enhances customer satisfaction by ensuring product availability. The results demonstrate how machine learning and data analytics can be effectively leveraged to provide actionable insights and support strategic decision-making in the competitive retail sector.

Keywords: Demand Forecasting, Machine Learning, Retail Analytics, Time Series Analysis, Prediction, Flask, Data Visualization.