Master of Science (MS)
First Committee Member
Number of Pages
In multiple linear regression involving several predictor variables, finding a suitable non-linear transformation of the predictors might be helpful to present the model in a simple functional form which is linear in the transformed variables. In this thesis, a computer code in C++ is developed to automate the process of finding a suitable transformation for the predictors. This is done by finding the transformation that yields the maximum correlation between the response and the transformed predictor. Several simulated examples are included to illustrate the method. A prime concern in calculating the correlation between two data sets is statistical accuracy. Correlation coefficients reveal the degree of correlation between two data sets. They are valued from -1 to 1. A positive value indicates correlation and negative values indicate anti-correlation.
Determination; Function; Linear; Multiple; Predictor; Regression; Transformation; Variables
University of Nevada, Las Vegas
If you are the rightful copyright holder of this dissertation or thesis and wish to have the full text removed from Digital Scholarship@UNLV, please submit a request to firstname.lastname@example.org and include clear identification of the work, preferably with URL.
Ravi, Vimatha, "Determination of transformation function for predictor variables in multiple linear regression" (2007). UNLV Retrospective Theses & Dissertations. 2136.
IN COPYRIGHT. For more information about this rights statement, please visit http://rightsstatements.org/vocab/InC/1.0/