Evolutionary Computation: History, Methods, and Applications

Explore the fascinating world of evolutionary computation in this comprehensive lecture presented by Josh Bongard from the University of Vermont at IPAM's Mathematics of Intelligences Tutorials. Delve into the history and main forms of evolutionary computation, tracing its roots back to the 1960s. Discover CPPN-NEAT, considered by many as the first generative AI method, and understand its continued relevance due to its ease of implementation and robustness to hyperparameter settings. Gain insights into practical applications through case studies from Bongard's research group, including symbolic regression, evolutionary robotics, and computationally designed organisms. Recorded on September 13, 2024, this 1-hour and 15-minute talk offers a valuable introduction to the field of evolutionary computation and its cutting-edge applications in artificial intelligence and robotics.

Josh Bongard - Evolutionary computation - IPAM at UCLA