Novel AI Model for Evaluating Buyers' Fulfilment with Clothing Fit

Abstract

Online merchants must make sure that their customers are satisfied with the fit of their products, yet standard size charts sometimes fall short of accounting for the wide variety of body types and preferences. This research presents a novel artificial intelligence (AI) model that aims to precisely assess buyers' fulfillment with clothing fit (BF-CF).  As test clothes, skirts composed of various textiles were utilized. The BF-CF was estimated by assigning predictor factors to the mechanical characteristics of the skirt textiles. Using a 3D body scanner, virtual body models of the study participants were produced and utilized for virtual fitting. Each participant evaluated how well the clothes fit after trying them on in person. Additionally, participants evaluated the virtual fit, which indicated how well they fulfilled the virtual fit (PVFF), by looking at the skirt simulations on their avatars. The data was warmed up to predict BF-CF using a novel machine learning technique called Tabu search optimized smooth linear logistic regression (TSO-SLLR). According to the experimental data, PVFF was the most important factor. In the fashion business, machine learning (ML) is mostly utilized for manufacturing and sales forecasting. Nevertheless, studies on the topic of clothing fit, which deters people from buying online, received little attention. Thus, we proposed an innovative machine learning approach in this work to forecast customers' fulfilment with clothing fit. This study shows how the AI model works to increase customer happiness, decrease returns, and improve the entire online shopping experience for both customers and clothes merchants through extensive testing and validation.