Engaging Undergraduate Students In Machine Learning Research: Progress, Experiences And Achievements Of Project Emd Mlr

Abstract

Abstract Project EMD-MLR is a National Science Foundation funded, on-going effort that aims at engaging undergraduate engineering and computer science students into research on Machine Learning. In our present paper we will provide a brief overview of the project’s characteristics and share our experience about engaging undergraduate students in research in year 1 of the project. More specifically, we will report the overall achievements accomplished so far in terms of research products such as student-developed software, publications and other dissemination efforts. Additionally, we report on student assessment results regarding the quality of their experience through their participation in aspects such as the student-teacher interaction, the knowledge and experience that students acquired, while performing research and the type of impact their involvement had on their future academic and/or career aspirations. 1. Introduction Machine Learning (ML) is a discipline that started evolving as early as the 60’s in the form of Artificial Intelligence and that nowadays has permeated several aspects of high-tech applications as well as everyday life. Its charter is to study, develop and build models able to perform “intelligent” tasks that may be second nature for humans, but are well beyond the capabilities of traditional computing paradigms. ML applications such as vending machines that recognize valid paper bills, document processing software that corrects our grammar and syntax in real time, voice-driven over-the-phone account management of credit, smart photographic cameras that automatically adjust their exposure and speed settings depending on the scene environment, as well as challenging strategy computer games, have become a big part of our everyday routine. Other, less obvious, applications, such as automatic target recognition, earthquake prediction, gene expression discovery, intelligent credit fraud protection and affectionate computing, to mention just a few, are examples of cutting-edge applications of ML in various technological, scientific and financial domains. This paper describes the outcomes of a prototype project titled “PROJECT EMD-MLR: Educational Materials Development through the Integration of Machine Learning Research into Senior Design Projects “, whose intellectual focus is ML. The project is an on-going, multi-institute effort that started in May 2004. The project partners are two major universities in Central Florida, namely Florida Institute of Technology (FIT) in Melbourne and the University of Central Florida (UCF) in Orlando. In addition to the two host universities, there are two 2-year Central Florida colleges, Seminole Community College (SCC) in Oviedo and Brevard Community College (BCC) in Melbourne. Project EMD-MLR is a National Science Foundation funded project under NSF grant CCLI-0341601 for the period of May 2003 to Arpil 2006 and under the auspices of the Educational Materials Development track of the Course, Curriculum and Laboratories Improvement (CCLI-EMD) program.