Title: ANALYZING SOIL FERTILITY USING DATA MINING TECHNIQUES

Authors: R. S. Parmar, V. I. Mehra and G. J. Kamani

Publisher: The President, Society of Extension Education Gujarat

Keywords: Data Mining, Soil Fertility, Yield Production, Classification, Agriculture

Volume: 34

Issue: 2

Year: December 2022

DOI: https://doi.org/10.56572/gjoee.2022.34.2.0011

Abstract: Data mining classification techniques play a significant role in agriculture. To increase yield production, many parameters are responsible like weather, fertilizers, pesticides, and soil. Soil is an important key aspect of agriculture, as an increase in yield can only be attained by ensuring that the soil provides a balanced and adequate supply of nutrients. The main aim of the work is to predict soil fertility using data mining classification techniques. Soil fertility is predicted using data mining classification techniques such as J48, Random Forest, Decision Table, PART, and Naive Bayes. These classifier algorithms are used to extract soil fertility knowledge from soil data. A comparison of different classifier algorithms concerning prediction accuracy revealed that the Random Forest is the best classifier to classify the soil fertility dataset. The Random Forest classifier algorithm can produce more consistent results of soil data for the Anand District of Gujarat State and the Kappa Statistics in the prediction were improved.