An innovative Deep Learning Based Approach for Accurate Agricultural Crop Price Prediction

Accurate predictions of crop yield and crop price provide valuable inputs for decision making by various stakeholders in agriculture: farmers, consumers, retailers, wholesalers, dealers, and the Government. Some of these decisions have far-reaching implications for the economic well-being of farmers, ensuring food security, stability of supplies, breeding of seeds, and for economic planning in general. This paper focuses on the problem of crop price prediction. There are several factors affecting crop prices. These include the expected yield, expected demand, export projections, import decisions, supply chain factors, weather conditions, geospatial factors, unanticipated events such as a pandemic or a flood, etc. Compounding this is the fact that the data that are available in many emerging economies about historical crop prices and crop price variations have several issues such as missing values, outliers, and even data entry errors. Accurate prediction of crop prices is therefore a grand challenge problem but at the same time an important one to help secure the economic prosperity of farmers. This paper focuses on how geospatial dependencies can be harnessed to obtain improved accuracy in predictions of crop prices.