A Comparative Analysis of Zinc-solubilizing Bacteria-mediated Chlorophyll Content in Wheat by Artificial Neural Networks and Random Forest
Nitesh Kumar Singh *
Department of Soil Science and Agricultural Chemistry, Institute of Agricultural Sciences, BHU, Varanasi, India.
Anand Prakash Singh
Department of Soil Science and Agricultural Chemistry, Institute of Agricultural Sciences, BHU, Varanasi, India.
Amit Kumar
Department of Agricultural Engineering, Narayan Institute of Agricultural Sciences, GNSU, Sasaram, Bihar, India.
Aditi Chourasia *
Department of Soil Science, College of Agriculture, RVSKVV Gwalior, Madhya Pradesh, India.
Shashank Shekher Singh
Department of Agronomy, Narayan Institute of Agricultural Sciences, GNSU, Sasaram, Bihar, India.
Tej Pratap
KVK Godda, Jharkhand, India.
*Author to whom correspondence should be addressed.
Abstract
This study compares Artificial Neural Network (ANN) models and Random Forest methods for predicting chlorophyll content in wheat under different zinc sources and bacterial inoculations. Experimental data from 2016–17 were analyzed. The total data set 32 were used and using four modeling approaches—two ANN architectures, Linear Regression, and Random Forest. Among them, Random Forest achieved the highest accuracy (R² = 0.83, RMSE = 1.50), outperforming ANN models due to better handling of small datasets. The results demonstrate that growth stage, zinc source, and bacterial treatment significantly influence chlorophyll levels. This study provides insights into selecting suitable modeling techniques for agricultural prediction tasks.
Keywords: Zinc solubilizing bacteria, zinc source, chlorophyll, ANN, RF, R2, RMSE, LR