Enhancing Feature Optimization for Crop Yield Prediction Models

Sabyasachi Chatterjee*, Swarup Kumar Mondal, Anupam Datta and Hritik Kumar Gupta 

Department of ECE, Heritage Institute of Technology, Kolkata, India

Corresponding Author E-mail: sabyasachi.chatterjee@heritageit.edu

DOI : http://dx.doi.org/10.12944/CARJ.12.2.19

Abstract:

The growth of the world population is leading to an increased demand for food production. Crop yield prediction models are vital for agricultural planning and decision-making, providing forecasts that can significantly impact resource management and food security. This paper focuses on the importance and benefits of feature optimization in enhancing the performance of crop yield prediction models. By reducing noise and complexity, optimized features allow the prediction models to concentrate on the critical factors affecting crop yield, leading to more precise predictions and lesser computation times. This work utilizes an enhanced genetic algorithm to optimize feature selection and model parameters, outperforming the performance of standard genetic algorithms. Comparative analysis showed significant improvement in the accuracy of yield predictions by optimizing the selection of relevant features. The minimal error between actual and predicted yields on both the training and testing datasets highlights the effectiveness of the enhanced genetic algorithm. Enhanced feature optimization not only improves the robustness and adaptability of yield prediction models but also contributes to more sustainable and efficient agricultural management.

Keywords:

Crop yield; Enhanced genetic algorithm; Feature selection; Genetic algorithm; Machine learning

Download this article as: 

[ HTML Full Text]