Major and Classification
Computer Science/Business Administration
Shrikanth Narayanan, Ph.D.
Viterbi: Electrical Engineering and Computer Science, Dornsife: Linguistics and Psychology
“Classifier Performance in Detecting Engaged and Disengaged Behavior in an Informal Learning Environment”
By using machine learning algorithms to detect engagement and dis- engagement in a classroom environment, the quality of instruction and content retention can be increased. Physical gestures provide a detailed behavioral analysis of the audience’s engagement to the instructor’s approach of teaching. This can aid the instructor to apply a different instructional technique that will fit the audience to the content. Automating the process of providing a thorough analysis of classroom engagement will be cost-effective and can potentially be done in real-time. Further research should be aimed towards the selection of features that will accurately capture audience member’s engagement and disengagement inferred from physical actions. Future applications of automated behavioral analysis can be applied in other observational settings such as diagnosing autistic children or couple’s therapy.