Prediction of Optimal Nearest Delivery Personnel for Delivery of Products Using Support Vector Machine Against Logistic Regression Algorithm for Better Accuracy

The goal of this research is to enhance the accuracy of online sales management by finding the most efficient logistics method from the manufacturer using machine learning techniques. Specifically, the study compares the effectiveness of Logistic Regression (LR) and Support Vector Machine (SVM) algorithms. The SVM classifier was tested on a dataset of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{8 0, 0 0 0}$</tex> records with 15 attributes, including a sample size of 125. This dataset, sourced from various web sources, focused on defective products. The study used a g-power value of 80 %, a significance threshold of 0.05 %, and a 95 % confidence interval. The research developed a framework to compare the effectiveness of SVM and LR algorithms for finding the best storage solutions in the supply chain. The results showed that the LR algorithm achieved an accuracy of 83 % in predicting defect rates, while the SVM algorithm achieved a higher accuracy of 95 %. The statistical analysis revealed a significant difference, with a p-value of 0.001, indicating that SVM performed significantly better than LR. This means that SVM is more effective in identifying the best warehouse for storing products, providing more accurate and reliable results compared to LR.