Suitable Crop Plantation Recommendation System for Profitable Harvest using Deep Learning

Abstract

Agriculture considered as one among the oldest of the occupation since Stone Age still draws many researchers towards it due to its inevitable nature and the necessity to feed the human being. Plantation and harvest are two major process in agriculture that requires many external factors to be considered. The elements which impacted the harvest planting designs are climatic conditions, yield forecast, and accessibility of assets, benefit and homestead zone. The market value of the agrarian product should match the cost incurred in production for an agrarian ware, if the market cost does not match with the cost of production that results in high debt or loss for the farmer. Examining the market estimation of a yield to be gathered at collect time is a critical choice for a farmer planning to benefit from his persistent effort. It proposes a recommendation system will enable farmers to predict the market value of a planted crop. The recommendation system uses Recurrent Neural Networks with Long Short-Term Memory (LSTM) cells to predict the price of the crop to be gathered in during the harvest.  The proposed system is analyzed with data from data website (www.data.gov.in) of the Government of India.