Evolution, Computation, and Learning

Evolution, Computation, and Learning
Instructor(s): 
Daniel Czegel

Semester: 

  • IAP 2019

Course Level: 

  • Graduate, Undergraduate
Class Days/Times: 
Mon 1:00pm to 3:00pm
Wed 1:00pm to 3:00pm
Fri 1:00pm to 3:00pm
Course Description: 
Add to Calendar Jan/14 Mon 01:00PM-03:00PM Location TBD
Add to Calendar Jan/16 Wed 01:00PM-03:00PM Location TBD
Add to Calendar Jan/18 Fri 01:00PM-03:00PM Location TBD

Enrollment: Limited: Advance sign-up required
Limited to 15 participants
Attendance: Participants welcome at individual sessions

Here, we will explore recent work in evolutionary computation and theoretical biology modeling the processes of evolution. Namely, we will focus on these broad questions:

  1. What are the processes that govern evolution or 'evolutionary learning'?
  2. How can these processes improve upon or inspire new models or theories of learning, search, and/or development?
  3. If any, what is the role of evolutionary computation or theoretical biology in investigating human cognition or developing AI?
  4. Are there any frameworks, theories, or models that we can import from these fields?

This course will include readings and ~30 minute lectures introducing general topics of interest such as evolutionary processes in the context of learning theory, what evolution can add to learning theory, evolvability and learning-to-learn, and complexification. The intention is to spark conversation about the role evolution plays in learning, how it can be further characterized or replicated in machines, and whether it is of interest or use to explore potential projects that build or expand on this recent work.

Sponsor(s): Brain and Cognitive Sciences
Contact: Felix Sosa, 305 733-6216, FSOSA@MIT.EDU