APPLICATION OF MATRIX IN VERSION IN MULTIVARIATE ANALYSIS

Abstract

Linear regression models are commonly employed to predict future response (dependent variable) values or to depict the effects of important explanatory variables on the responses. Unfortunately in most regress problems, data were routinely recorded rather than generated form pre-selected settings of the independent variables (the designed experiment) and in these cases, the independent variable were frequently linearly dependent. This paper discusses how matrix inversion can be used for computational purposes and  for interpreting multivariate data.