Creating predictive clothing size models for online customers. Issue 4 (21st April 2023)

• Record Type: Journal Article
• Authors: Davidson, Allison; Gundlach, Ellen
• Is Part Of: International journal of mathematical education in science and technology. Volume 54:Issue 4(2023)
• Page Start: 614
• Page End: 629
• Publication Date: 2023-04-21
• Subjects: Linear regression -- multivariate analysis -- prediction intervals -- online shopping; Mathematics -- Study and teaching -- Periodicals; Mathematics -- Periodicals; 510.07
• DOI: 10.1080/0020739X.2022.2040623
• Languages: English
• ISSNs: 0020-739X

Abstract

Abstract : A disadvantage to online clothes shopping is the inability to try on clothing to test the fit. A class project is discussed where students consult with the CEO of an online mensware clothing company to explore ways in which an online clothing customer can be assured of a superior fit by developing statistical models based on a shopper's height and weight to predict measurements needed to create a suit that feels custom-made. The dataset is most amenable to use with students who have previously been exposed to simple linear regression, and can be used to explore multiple regression topics such as interaction terms, influential points, transformations, and polynomial predictors. Discussion points are included for more advanced topics such as canonical correlation, clustering, and dimension reduction.