Study on Establishing Regression Model to Predict Leg Girth of Young Women

Abstract

Using Martin Meter from Japan to measure 110 young women body. Using SPSS software for correlation analysis of measurement data, then selecting width, thickness and weight as elements used for establishing leg girth regression models. Adopting backward regression method to choose variable further, and establishing Multiple Linear regression model, Multivariate Quadratic regression model, Multivariate Cubic regression model to predict the girth of each part of leg which related to clothing. After testing and comparing the effect of each model, choosing out the best fitting modle for each part. The validity of modles were tested by randomly selecting 20 young women’s leg data,comparing with manual measurements, gaining satisfactory results. So as to provide an important technical support for human leg two-dimensional non-contact measurement and be further applied to research on medical compression stockings.