International Journal of Science, Engineering and TechnologyAuthors: Sivaranjani S, Kanishka B, Mr Sankaran D
Abstract: Thus, agricultural price volatility is found to affect farmers’ income, agricultural markets, and policy planning. Price forecasting helps agricultural producers make informed decisions. In this context, the paper proposes a SmartFarm Price Advisor framework that utilizes machine learning techniques to accurately forecast commodity prices based on past modal, minimum, maximum prices, and rate of change. This paper analyzes commodity prices for a multi-year time frame, preprocesses the data to obtain extensive insights, and produces exploratory results like commodity price trends, statistical summaries of individual commodities, and correlation structures between variables. Various supervised learning and time series techniques are implemented and studied to find the most appropriate methodology suitable for forecasting commodity prices for different crop categories. From the experimental implementation, the correlations between key commodity prices are found to be strong, while significant commodity-specific changes are observed. Therefore, the proposed commodity price system has great potential to be implemented in agricultural settings.
