Forecasting of Tomato prices and yield based on season and weather using LSTM and Arima algorithms

The tomato price volatility in India is a cost to the farmers and consumers alike in time resource allocation and in budgeting. 2. In order to overcome these failings this research develops a predictive framework based on machine learning algorithms (i.e., LSTM, ARIMA) to predict the price of tomato. An analysis of their historical prices, seasonal trends and weather conditions provides us not only the basis of their accuracy in predicting prices but also the dynamism of their predictions. Compared to LSTM models, ARIMA is statistical and for better prediction accuracy while the later models the temporal structures in prediction use complex temporal structures. The proposed system provides these stakeholders with the opportunity to better inform decisions on price fluctuations through its real time intelligence. Timely forecasts are sought by the framework to reduce financial uncertainties and optimize resource usage along the agricultural supply chain. A final agricultural market management research question that finally provides for the more efficient management of agriculture markets through the ability of the predictive model to predict the changes in price.