Regression Analysis on Geometric Transform and Applications

Abstract

Regression analysis is one of the popular statistical tools to construct a predictive model of a scalar variable. For constructing a predictive model of an event in 2D/3D space, a 2D/3D vector is used to represent the event and the predictive models are constructed for each of the elements of the vector inde- pendently. In this paper, we focus on geometric transforms in 2D/3D space such as rotation and translation and formulate regression analysis on them using geometric algebra. We demonstrate the descriptive power of the model using some applications.