Paper Title
Price Prediction of Jute For Five Districts of West Bengal Using Deep Learning Technique

Abstract
Abstract - Jute is referred to as 'Golden Fiber' not just because it appears shiny and yellow but also because it is a good source of livelihood and income for many indigenous farmers and industrial workers. India is the leading jute-producing country globally, and West Bengal is the largest producer of jute among Indian states. It is one of the most cultivated natural fibers in West Bengal, where it supports innumerable families by generating plenty of employment opportunities and earning valuable foreign exchange. Unfortunately, jute prices are vulnerable to fluctuations caused by weather, market conditions, and several other factors.When they cannot sell their produce and earn a profit, farmers and traders have to bear the brunt of the changing market prices. Hence, it becomes essential to know and to be able to predict the wholesale price of a crop like jute. This study aims to predict the price of jute for the districtsof CoochBehar, Malda, Nadia, Uttar Dinajpur, and Murshidabad of West Bengal. A recurrent neural network - Long-Short Term Memory is used for this purpose. The time-series data of eighteen years was collected for this study starting from January 2002 up to December 2019. Keywords- Deep Learning, Long-Short Term Memory, RMSE, MSE, MAE