Higher Accuracy of Drought Prediction Using Linear Discriminant Analysis Algorithm Over AdaBoost Algorithm

This study focuses on improving drought prediction using two machine learning algorithms: Linear Discriminant Analysis (LDA) and AdaBoost. The research involved classifying data into six levels of drought using 18 meteorological indicators over a period of 90 days, with each algorithm tested across 10 iterations. The G-power test was set at 80%. Results showed that the LDA algorithm achieved a mean accuracy of 94.82%, while the AdaBoost algorithm reached 79.66%. An independent t-test indicated a statistically significant difference between the two algorithms, with a p-value of 0.001, which is less than 0.005. This study aims to enhance drought prediction efficiency and accuracy by comparing these algorithms. The findings clearly demonstrate that the Linear Discriminant Analysis Algorithm outperforms the AdaBoost Algorithm in terms of accuracy and effectiveness.