Enhancing predictive accuracy: Statistical approaches to address multicollinearity in agricultural studies

Multicollinearity poses a challenge in regression analysis, leading to unstable estimates of regression coefficients and complicating the interpretation of explanatory variables. This study addressed the issue of multicollinearity in the context of multiple linear regression (MLR) using principal components regression. The findings underscored the efficacy of principal component regression in addressing multicollinearity using performance criteria with R² of 0.946, RMSE of0.576, AIC of 21.227 and BIC of 23.818.