Soil Based Prediction for Crop Yield using Predictive Analytics

Abstract

Soil is the main component and plays a significant role in agriculture. Based on the nutrients and pH value of the soil, crop yielding is determined. Farmers are still using traditional approach to analysis the soil quality. The techniques like Data Mining, Artificial Intelligence, Machine Learning, Deep learning and Predictive Analytics are the emerging technologies in research to improve the agricultural field. Predictive analysis is a technique of machine learning that predicts the future outcomes and analysis is based on the historical or past data. In agriculture, predictive analytics helps to predict or identify the soil nutrients level required for the crops like Paddy, Raagi, Cumbu etc.,. In this paper, the soil based dataset is collected from TNAU website and it has 32 districts of Tamilnadu. The algorithms such as Naïve bayes, Bayes Net, and IbK have been deployed to predict the crop variety suitable for the soil based on the total production and area sown district wise. Also, its accuracy levels are compared. The accuracy is determined using true positive value, false positive value, precision, recall, f-measure and MCC.