Implementation of Multilayer Neural Network with Decision Tree Model for Classification of Soil Type and suggesting suitable Crop Cultivation using Machine Learning Technique

Abstract

Soil is a significant element in agriculture. Several types of soil exist in different parts of our Country. Each type of soil has its own characteristic features and support growth of different kinds of crops. The yield of agriculture purely depends on environmental conditions and soil type. We need to know the features and characteristics of various soil types to understand which crop grows better in certain soil type. For the purpose of finding the classification of soil types the soil dataset is downloaded and based on the soil type predicted the farmers are suggested to cultivate the suitable crop. In this research work the various soil types are classified with the help of proposed algorithm using Multilayer Neural Networks with Decision Tree model. Experimental results illustrates the performance of generating best decision tree for classifying soil type from the given soil dataset. The algorithm for application of Multilayer neural networks with Decision Tree model helps to classify the soil types more accurately than the existing algorithms such as  SVM, KNN, Bayesian approach to decision tree and Ensemble approach to Decision Tree model.