Enhancing Clothing Size Prediction in E-Commerce: A Machine Learning Approach for Improved Customer Experience

Clothing and apparel companies face high rates of returns due to customers ordering incorrect sizes. This can be frustrating for customers and is a waste of time and resources for the company, as they have to process the return and potentially restock the item. Uncertainty about fit also deters online purchases, impacting sales. Most currently available systems can predict the size based on all measurements provided by the customer, but that is not always feasible. This brings a need for the integration of Machine Learning to solve this problem of predicting sizes and provide better results with increased accuracy. This paper presents a thorough evaluation of various ML models for size prediction, namely Logistic Regression, K-Nearest Neighbors (KNN), Random Forest, and Decision Tree models. The results indicate that our model showcases considerable promise for enhancing sizing accuracy, with the Random Forest and Decision Tree models achieving noteworthy accuracy scores of 0.99 in shirt size prediction. In the realm of pant size prediction, these models maintain their effectiveness, with an accuracy score of 0.89 and 0.82, respectively. This underscores the potential of our ML model to alleviate sizing woes in the clothing retail industry.