Using Machine Learning to Predict and Promote Student Success

Concurrent Session 9

Brief Abstract

Machine Learning has become ubiquitous in the technology world. Can the same technology powering search engines and shopping recommendations help instructors in the classroom? Is it ethical? We will provide an overview of the machine learning research and development process and show some real world examples of schools using machine learning algorithms to promote student success.

Presenters

Lead Presenter: David Cramer, StrongMind

David Cramer is a software developer at StrongMind, a full service curriculum/SIS/LMS provider for online schools, focusing on implementations of machine learning and predictive analytics. In previous capacities at Strongmind he has developed reporting and business intelligence solutions. Prior to working at StrongMind, he was the primary student data analyst at Primavera Online High School - the largest school in the state of Arizona. In that capacity he was responsible for managing the reporting and data needs for the entire school with a particular focus on contributing to the growth and effectiveness of PLCs. He also oversaw the implementation of third party data and benchmarking tools such as Data Director and conducted school-wide training promoting the use of data in the digital classroom.

Additional Authors

Jason Tourville, Primavera Online Highschool

Jason Tourville is currently the Assessment Coordinator at Primavera HS / MS, the largest online school in Arizona. His duties as an Assessment Coordinator are to assist educators in understanding the purpose of data driven instruction through analyzing data dashboards and applying interventions to student data reports. He has received his B.S. in Mathematics from Arizona State University, M.Ed. in Mathematics Curriculum from Plymouth State University and currently is taking the fourth in the series of the entry level ASA actuary exams and pursuing a Masters in Applied Statistics from the University of Kentucky.

Janardhan Manjunath

Extended Abstract

Desired Outcomes:
By the end of the presentation all attendees should:
1. Be familiar with the general principals of machine learning and its current use in technology and education. 
2. Have an awareness of the process of machine learning feature selection, the ethical implications and concerns of different feature domains, and a general notion of the time and understanding required for data cleaning.
3. Have knowledge of machine learning use cases currently employed by schools, their effectiveness, and areas for their improvement and/or expansion.