Crop Yield Prediction Using Random Forest Algorithm for Major Cities in Maharashtra State

Abstract

Maharashtra is a leading State in  agriculture. Agriculture is the one that plays important role  in the economy of India. India is an agricultural country and  its economy largely based upon crop production. Hence one  must say that agriculture is often the backbone of all  businesses in the a-part-of-us country. Basically paper  focuses on predicting the yield of the crop by using a  different machine learning algorithm. The application  (Smart Farm) developed in this research helps users to  predict the crop yield using different climatic parameters.  Different methods of predicting crop yield are developed  within several years with different outcomes of success, but  many of them do not take into consideration the climate.  Machine Learning is the best technique which gives a better  practical solution to crop yield problem. So the Random  Forest algorithm which we decided to used to train our  model to give high accuracy and best prediction., we chose  5 climatic parameters to train the model. Agriculture  inputs such as pesticides, fertilizers, chemicals, soil quality,  etc. were not used as it depends upon the type of field. The  model is trained and designed using 20 decision trees build  the random forest algorithm which gives better accuracy of  the model. 10-fold cross-validation technique used to  improve the accuracy of the model. The predicted accuracy  of the model is analyzed 87%.