<![CDATA[In the era of customer-centric logistics, the ability to predict delivery time preferences is critical to enhancing service quality and operational efficiecy. This study proposes a predictive analytics framework to classify customer-preferred delivery times – morning, afternoon, evening, or night – using historical logistics data and customer perception analysis. Utilizing the Amazon delivery dataset, categorical time intervals were derived through timestamp transformation, enabling classification modeling with three machine learning algorithms Naive bayes, Logistic regression, and Decision tree. Among these, the Decision tree algorithm yielded the best performance, achieving an accuracy of 77% and a macro F1-score of 0.77. Further analysis revealed that traffic conditions, vehicle types, and product categories were the most influential features in predicting delivery time preference. Survei results corroborated the model’s findings, with customer responses highlighting traffic and delivery timing as top priorities for service satisfaction. This research demonstrates the integration of data-driven modeling with customer insights to support decision-making in last-mile logistics. The findings can guide logistics providers in designing adaptive, preference-based delivery schedules that improve service realiability and user experience.Keywords: predictive analytics, delivery time preference, decision tree, last-mile logistics, customer satisfaction, machine learning.]]>
Copyrights © 2026
