Appraising relation between weather and Red gram yield estimates with
Regression Machine Learning

The frequent surge in the price of red gram as compared with other pulses necessitated seed yield estimation to cope up with demand- supply equilibrium by policy-makers and ef icient resource utilization by farmers and agronomists. Besides, to achieve another main objective of appraising the relationship between red gram crop yields and weather, ivesupervised regression machine learning algorithms namely,Gaussian processes, Linear regression, Support vector machines, k-Nearest neighbors and Decision tree were used 2 in the study. Among these tested algorithms, the Random Forest algorithm was better with crop yield predictability of 95 % (R ), lowest Mean Absolute Error (MAE) of 32.7 and Root Mean Squared Error (RMSE) of 40.8 as compared with other itted regression algorithms. It was further noticed that, the actual yield and the predicted yield based on training data set were close to each other and the residual ranged from -76 to 99, while it ranged from -148 to 111 in case of testing data by the same Random Forest model.