A DATA MINING APPROACH TO CROP YIELD PREDICTION USING MACHINE LEARNING
One of the essential industrial sectors in India is Agriculture and the economy of a country is relied on it highly for sustainability in rural areas. The Indian agriculture level reduces gradually owing to some factors like excessive utilization of pesticides, water level decrement, climate changes, and unpredicted rainfall, etc. On the agriculture data, descriptive analysis is performed to understand the production level. Because of lack of ecosystem control technologies deployment, most of the agriculture fields are under developed. The production of crops is not increased owing to these issues that affects the economy of agriculture. Based on the prediction of plant yield, agricultural productivity is improved. By using machine learning techniques, the crop from given dataset have to predict by agricultural sectors for preventing this issue. To capture the information, the dataset analyses by supervised machine learning technique (SMLT). The effectiveness of proposed method of machine learning algorithm can compare with best accuracy based on the results. In this paper use ANN with cascade-forward backpropagation and Elman backpropagation for yield prediction. To determine input variables that maximize the interested neurons’ activation, the positive gradients backpropagate by Cascade-Forward backpropagation method. A recurrent connection exists from the hidden layer output to its input is included in Elman backpropagation network which is a two layer backpropagation network. These two techniques are better prediction techniques compared to ANN with backpropagation.