ML-Based Insights for Crop Yield Forecasting in Indian Farming

Nitish Kumar1 Dr. Pankaj Richhariya2
  • 1 Research Scholar, Department of Computer Science, Bhopal Institute of Technology & Science.
  • 2Hod, Department of Computer Science, Bhopal Institute of Technology & Science.


This study investigates the use of machine learning methods to forecast crop yields in Indian agriculture, using a large dataset that covers the period from 1997 to 2020. Data preparation methods, such as feature selection, one-hot encoding, and transformation, were used to improve the quality and appropriateness of the dataset for modelling. Individual regression models, including Decision Tree, Random Forest, Support Vector Machine (SVR), and K-Nearest Neighbours (KNN), were trained and assessed. These models showed strong performance in accurately forecasting crop yields. Furthermore, the integration of a hybrid ensemble model, which incorporates voting ensemble approaches, resulted in higher predicted accuracy and increased model resilience. Our technique stands out from previous approaches due to its uniqueness and improvements. These include the use of a bigger dataset, a longer time frame, and the use of hybrid ensemble models. The results provide useful insights for stakeholders engaged in agricultural decision-making, enabling informed allocation of resources and implementation of risk management measures to improve productivity and sustainability in Indian agriculture.

Vol 3 Issue 2 (2024)

Journal Details

   ISSN : 2583 – 7117

Most read last week

Organisation Website