WSEAS Transactions on Advances in Engineering Education
Print ISSN: 1790-1979, E-ISSN: 2224-3410Volume 16, 2019
Pedagogy of Bootstrapping
Author:
Abstract: Demos created to illustrate the bootstrapping procedure and its application to machine vision are presented with special emphasis on the area under the receiver operating characteristics (ROC) curves associated with a sensor. These formed an integral part of the undergraduate course in engineering probability updated to include topics in data analytics. Starting from the basics, the demos offer a step-by-step way to implement bootstrapping and statistical analysis of the areas under the ROC curves. A smaller data set was initially used to articulate all the details before analyzing the larger data set. Students were able to follow the procedures alongside. Results were also verified contemporaneously using commercial software used in ROC analysis. Furthermore, the efficacy of the demos and its effect on the learning process was evaluated through a set of individual assignments to the students followed by the statistical analysis of student surveys given at the beginning and conclusion of the course. They appear to show enhanced understanding of bootstrapping and ROC analysis by the students. The demos and the methodology proposed here could easily be extended to cover other topics of interest to expose the students to didactic aspects of complex concepts in engineering probability.
Pages: 18-27
WSEAS Transactions on Advances in Engineering Education, ISSN / E-ISSN: 1790-1979 / 2224-3410, Volume 16, 2019, Art. #3