IMLAPC: Interfused Machine Learning Approach for Prediction of Crops

Abstract

Agriculture is the main occupation of rural India, which promotes economic growth in the country's development. To increase the yield of the crops to feed the increasing population, it is essential to identify the crops which can be grown in the respective zones. In this article, the Fused Classifier Algorithm (FCA) and Interfused Machine Learning Algorithm (IMLA) are proposed to predict crops suitable for the land based on the zones and agro-climatic parameters. Focusing on the zones of the Karnataka region, the model predicts the crop to the farmers. The different machine learning models such as naïve Bayes, decision tree, neighbors, multilayer perceptron have also been evaluated by varying the hyperparameters and checked for accuracy of the models built. The FCA algorithm merges different algorithms using the error rate with hyperparameters tuning and is given to IMLA to predict crops. This article also compares different machine learning classifiers with the proposed IMLA algorithm, which shows better accuracy with 82.7%.