Enhanced Forecasting Methods for Best Route Analysis by Random Forest Algorithm Over Naive Bayes Algorithm

This study uses a publicly accessible dataset from Kaggle to evaluate the performance of a novel Random Forest technique for optimal route analysis against a Naive Bayes classifier. The evaluation involved a sample size of 20, split evenly into two groups (Group 1 and Group 2, each comprising 10 samples), and calculations were conducted using G-power with parameters set at an alpha of 0.05 and a beta of 0.2, corresponding to a 95% confidence interval. The route analysis was conducted using the RF method with 10 samples and the Naive Bayes (NB) classifier with 10 samples as well. The results indicate that the novel Random Forest algorithm achieved an accuracy rate of 95.00%, significantly outperforming the Naive Bayes classifier, which achieved an accuracy rate of 86.33%. The statistical significance of the findings is confirmed with a p-value of 0.030, which is less than the 0.05 threshold. This study demonstrates that the proposed Random Forest technique significantly surpasses the Naive Bayes classifier in terms of accuracy for route prediction tasks.