Rice yield prediction in Dakshina Kannada district using ensemble machine learning

Agriculture is the primary driver of Indian economy and country's GDP growth.Forecasting of agricultural production is an essential element for efficient resource management.Rice is primary crop grown in Dakshina Kannada District for three distinct seasons, each having its own set of climatic conditions.Furthermore, this region receives maximum rainfall and experiences fluctuations in monsoon intensity.This generates complex models in predicting rice crop yield.The machine learning (ML) techniques have proven effective in modelling using both historical and current data to predict crop yield (Aljuaydi and Wu 2022).ML algorithms like Random Forest (RF), Support Vector Machines (SVM), k-Nearest Neighbours (kNN) and Artificial Neural Networks (ANN) are used to predict crop growth (Shawon et al., 2024).As example, ML models can analyze hidden relationships between soil health parameters and climatic variables (Zou and Okhrin 2024).Development of crop yield models for seasons-based agriculture is crucial (Paudel, 2020).In rabi season, the crop exhibits stability during its growth, while in Kharif seasons, the yields become erratic due to monsoon fluctuations (Mohapatra et al., 2023).