Title: The Precursor of Theory for Discrete EAs: What Happened 25 Years Ago?

Speaker: Martin Krejca

Abstract: 

Run time analysis of evolutionary algorithms (EAs), that is, proving
mathematical guarantees about the run time of EAs, is currently one of
the major areas in the theoretical domain of evolutionary computing. By now,
the run time of certain EAs on typical benchmark functions is very well
understood, ranging from upper and lower bounds on the expected run time
over calculating leading constants to even having a thorough understanding
of the entire distribution of the run time in specific cases. But was this
always the case?

In this talk, we take a look at what theoretical analyses of discrete EAs
looked like 25 years ago, when this field was still in its infancy. We dive
into the PhD thesis of Günter Rudolph, and we discover what aspects were
relevant back then, what results could be derived, what tools were used, and
how this compares to our current knowledge. Buckle up for some interesting
(re)discoveries!