Piecewise regression for determining Malaysian woman foot shape

Abstract

A novel procedure to determine foot shape is propose. The relationship between shape features is used to discriminate shape. Five foot shape features that are Left Foot Length (LFL), Left Foot Breadth (LFB), Left Ball Girth Circumference (LBGC), Left Instep Length (LIL) and Left Fibulare Instep Length (LFIL) from a random sample of 161 Malaysians ladies were studied. Two groups of foot shape feature obtained from standard clustering procedure were then used to discriminate foot shape using piecewise regression. A two component piecewise regression model using LFB and LBGC give the best fit to the data showing two categories of foot shape.