Drought Assessment Accuracy Prediction Using Artificial Neural Network and Random Forest for Anantapur District

This study's overarching goal is to evaluate two drought prediction approaches using ANN and RF, respectively. Tools and Techniques for the Research: For the purpose of drought assessment performance prediction and analysis, two groups were established, one using an Artificial Neural Network technique and the other using a Random Forest approach. A total of 134 samples were collected, with 67 samples allocated to each group (varying proportions for training and testing). With a 95% confidence range, the Gpower test yields around 85% for
$\alpha= 0.05$
and
$\text{power} =0.85$
. The results show that compared to Random Forest (89.9560%), Artificial Neural Network (92.6420%) obtains significantly greater accuracy (
$\mathrm{p} < 0.05$
). Artificial Neural Network is obviously superior to Random Forest when the two methods are compared side by side.