Fig. 5

Backward stepwise elimination regression based on Akaike Information Criterion (AIC).This plot illustrates the backward stepwise elimination process used to refine the model by sequentially removing variables to minimize the AIC. Each point represents a step in the elimination process, showing the remaining variables and corresponding AIC values. The initial model, with an AIC of 74.482, is reduced to a final model with an AIC of 60.540. The process identifies the most parsimonious model by retaining only the variables that contribute significantly to model performance, optimizing predictive accuracy and interpretability.