Performance Evaluation of Supervised Machine Learning Algorithms for Customer Classification in E-Commerce

Customer classification plays a pivotal role in enhancing personalization, marketing strategies, and overall decision-making in E-Commerce platforms. This study aims to evaluate and compare the performance of several supervised Machine Learning (ML) algorithms for customer classification tasks. Specifically, we investigate the efficacy of Support Vector Machines (SVM), Decision Trees (DT), K-Nearest Neighbors (KNN), Random Forest (RF), and Logistic Regression (LR) using a real-world customer dataset collected from an E-Commerce platform. The dataset comprises multiple behavioral, transactional, and demographic features. Models are trained and tested using stratified ۱۰-fold cross-validation to ensure robustness. Performance is measured through standard classification metrics, including Accuracy, Precision, Recall, F۱-Score, and AUC-ROC. Experimental results reveal that ensemble-based methods such as RF outperform individual learners in most scenarios, particularly when dealing with noisy or imbalanced data. In contrast, algorithms like SVM and LR demonstrate competitive performance on linearly separable datasets with fewer features. The findings offer valuable insights for practitioners and researchers aiming to implement effective customer classification systems in dynamic online retail environments.

کلیدواژه ها:

Customer Classification ، Supervised Machine Learning ، E-Commerce ، Random Forest ، Support Vector Machine ، Performance Evaluation