Title: Improving real-world learning using scalable automated teachers
Date/Time: August 18, Thursday, 11:00 AM – 12:00 PM EST.
Abstract:
What are the most effective ways to teach, and how quickly can humans learn? To begin to answer these questions, it can be instructive to think back to your personal learning journey. Who has been your most effective teacher? What was your favorite thing to learn about? Which concepts or skills were more difficult for you to learn, and how did you overcome those difficulties? Broadly, the answers to these questions likely depend on what specific content is being learned (e.g., introductory coding in Python versus medieval European history), properties of the individual learner (e.g., a kindergartener versus a graduate student), and myriad other factors. My lab is working to build scalable general-purpose solutions to the "teaching problem"-- i.e., teaching any person, any content, as quickly and effectively as possible. Our approaches to optimizing instruction center on characterizing (a) the knowledge and skills the learner has already acquired, (b) the set of concepts the learner still needs to know about or master, and (c) the optimal timings and formats for presenting new material. In addition to drawing on my own experiences as an educator, this line of research incorporates advances in cognitive modeling, natural language processing, natural language understanding, human-computer interaction, and other related fields. I'll talk about some of our prior and ongoing work on developing automated teachers that can be scaled to support millions of students and thousands of content areas.
Bio of the speaker:
Jeremy Manning is an assistant professor of Psychological and Brain Sciences at Dartmouth College and directs the Contextual Dynamics Laboratory. His research is on understanding the brain network dynamics that support real-world learning and memory. He holds BS degrees from Brandeis University in Computer Science and Neuroscience, and a PhD in Neuroscience from the University of Pennsylvania. He completed his postdoctoral training at Princeton University. Professor Manning is passionate about computational training and methods development. He leads the development of several popular Python toolboxes including hypertools, supereeg, and timecorr, and is a contributor to several others. He Co-Directs the Methods in Neuroscience at Dartmouth Computational Summer School and teaches several open courses on data science and cognitive neuroscience.
This webinar will be offered online via Zoom. Please register to receive the Zoom link prior to the webinar.
------------------------------
Jianjun Hua
Statistical Consultant
Dartmouth College
------------------------------