Paper Title
DEVELOPMENT OF AI-ML BASED MODELS FOR PREDICTING PRICES OF AGRI-HORTICULTURAL COMMODITIES

Abstract
The Department of Consumer Affairs (DoCA) monitors daily prices of 22 essential food commodities across India through a network of 550 price reporting centers. To stabilize market prices, the government maintains and strategically releases buffer stocks of pulses and onions. Current forecasting methods rely predominantly on ARIMA-based econometric models that utilize seasonality, historical price trends, and market intelligence. However, these traditional approaches have limited capacity to capture non-linear patterns and respond to sudden market fluctuations. This research presents the development of AI and machine learning-based predictive models for forecasting prices of pulses and vegetables, with particular emphasis on onion and potato commodities. By integrating multiple data sources—including historical price data, meteorological information, and agricultural production statistics—the proposed system aims to deliver more accurate and timely forecasts. The outcomes of this study will assist government agencies in making proactive, data-driven decisions for price stabilization and effective buffer stock management. Keywords- Agricultural commodity forecasting, price prediction, machine learning, SARIMAX, LSTM, XGBoost, buffer stock management, policy decision support