Quality Control of Body Measurement Data Using Linear Regression Methods

Determining the accuracy of measurements for different types of objects is a complex task because (1) allowable deviations depend on the intended use of the measurements, and (2) different applications use different methods to identify incorrect measurements. In the construction of clothing patterns, up to 50 different human body measurements are used, and each person is unique. Therefore, it is difficult to draw conclusions from a specific set of measurements as to whether the included measurements are correct or not. In this study, the quality of body measurement data is being verified using linear regression methods. Linear regression models are trained with 80\% of the entire dataset, while the remaining 20\% is used for testing. The obtained results have been validated on a real dataset, and they allow for predicting missing or inaccurate body measurements with sufficiently high accuracy.

---

1. Thematic Sessions: Regular Papers

---

2. Opracowanie rekordu ze środków MNiSW, umowa nr POPUL/SP/0154/2024/02 w ramach programu "Społeczna odpowiedzialność nauki II" - moduł: Popularyzacja nauki (2025).