Machine Learning and Pattern Recognition Education at a non-Ph.D. Granting Engineering Program – Opportunities and Challenges

It is not very often that an institution gets the opportunity to design an engineering curriculum from scratch; an opportunity that also comes with many challenges, as well as rewards. Apart from how to present which material in the most innovative, efficient and pedagogical way, there is also the ubiquitously delicate issue of how to balance teaching and research – a significant challenge considering that this new program is housed at a primarily undergraduate institution with no Ph.D. component. We have therefore formed a number of focus areas based on the faculty interest, one of which is machine learning and pattern recognition. We have developed a number of courses, all of which are offered to senior undergraduates as well as M.S. level graduate students. Lacking a Ph.D. program, we also had to find creative ways of integrating research activities into the undergraduate and M.S. level curricula, which we have done so through the “Engineering Clinic.” The hallmark of Rowan’s engineering program, the engineering clinic is an 8-semester sequence, where vertically and horizontally integrated students work on externally funded semester-long projects for credit. In this paper, we describe our unique approach to integrating teaching and research activities under the engineering clinic umbrella, and how this approach allows us to offer research experiences even to under classmen. We also describe the specific opportunities and difficulties this approach brings, and we describe recently awarded grants on incremental learning and data fusion as examples of how we maximize these opportunities and how we address the difficulties.