A Supervised Machine Learning Regression Approach for Yield Prediction of Summer Rice of North-Eastern Region of Assam, India
Pal Deka *
Department of Agricultural Statistics, Biswanath College of Agriculture, Biswanath, Assam, India.
Pranjal Kr. Chakravartty
New Mangalore Port Authority, Mangalore under MoPSW, India.
Bhaskarjyoti Talukdar
Department of Statistics, Jhanji Hemnath Sarma College, Jhanji, Sivsagar, Assam, India.
Prateeti Bora
Department of Statistics, Gauhati University, India.
Deep Prasad Panika
Nilachal College of Allied Health Sciences, Gauahati, Assam, India.
Kakumoni Borah
Spectrum Academy, Madhabpur, Narayanpur, Assam, India.
*Author to whom correspondence should be addressed.
Abstract
Machine learning applications in predicting rice varieties in Assam have shown significant promise, leveraging various algorithms and data sources to enhance accuracy and efficiency. These applications range from predicting rice production to classifying rice varieties using advanced machine learning techniques. The purpose of this study was to predict the yield of summer rice varieties grown in Assam's northeast between 2007 and 2015. The R-square value for Bayesian ridge (0.97) and Lasso regression (0.96) was found to be high followed by Random forest algorithm (0.94). The decision tree algorithm (0.87) gives least score for variation explained (R-square) with respect to rest of the algorithms. From the evaluation metrics, mean absolute, median absolute and root mean errors for Bayesian ridge algorithm were less among the other regression algorithms.
Keywords: Summer rice, Prediction, Regression algorithms, North-east, Modelling