Informatics skills, which include machine learning, data science, and data mining, are rapidly becoming extremely valuable in science and engineering research. The accessibility and low cost of powerful informatics tools, the relative simplicity of the research problem framing, and the virtual nature of the work means that students with limited background knowledge can perform impactful research with some basic training that requires limited guidance and funds. Research can also be enhanced by the inexpensive and scalable cyberinfrastructure currently available (e.g., online databases, open-source codes, social networking tools). The Informatics Skunkworks program, initially developed at UW-Madison, includes training materials and professional development activities designed to provide undergraduates the skills needed to solve research questions using applied informatics. The Informatics Skunkworks training materials are designed to enable researchers to leverage applied informatics to engage undergraduate teams in solving authentic research questions within one semester. During the training, students work through educational materials and problems using real data sets to create machine learning models, assess model performance, explore cross validation strategies, and optimize model hyperparameters. Throughout the program, students also work on professional development activities designed to help them develop an understanding of what research is and how to effectively engage in the process. These critical research skills include effective record keeping, establishing group expectations, networking with other professionals, written and verbal communication. Because the Informatics Skunkworks program is primarily virtual and requires no laboratory space or equipment beyond a computer, our goal is to eventually enable students and faculty from institutions with limited research programs and resources to engage in authentic research projects that utilize informatics tools to answer important questions.

Using machine learning to engage undergraduate students in authentic, accessible and low-cost research projects

ACS Spring Meeting 2021

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.