GBEA Workshop at GECCO 19

Workshop Schedule

Sunday, July 14, 16:10-18:00, South Hall 1A (1F)

  • 16:10-16:40 Game Benchmark for Evolutionary Algorithms: An Overview
    Vanessa Volz, Tea Tusar, Boris Naujoks, Pascal Kerschke
  • 16:40-16:55 Game AI Hyperparameter Tuning in Rinascimento
    Paper: Ivan Bravi, Vanessa Volz, Simon Lucas
  • 16:55-17:45 How can EC and games researchers learn from game benchmarking results?
    Panel: Dimo Brockhoff, Jonathan Fieldsend, Mike Preuss, Simon Lucas
  • 17:45-18:00 General discussion on (real-world) benchmarking
    Everyone

Benchmarking SIG Meeting

Sunday, July 14, 20:00-21:00, GECCO Reception

Are you interested in benchmarking? Join our discussion!

Seeing how many workshops around benchmarking are taking place at GECCO and beyond, there seems to be lots of interest in the topic but not much concerted effort. We think it might make sense to informally meet up with everyone who is interested in the topic and related ones. We will thus meet up at GECCO to collect open questions and comman issues and plan some further steps to organise joined efforts in benchmarking, such as seminars.

See our agenda and updates here!

In the long run, we want to discuss questions such as:

  • What are the similarities and differences between different projects / workshops etc. around benchmarking EAs?
  • What are existing benchmarking frameworks, what are the differences and could / should they be merged?
  • What are appropriate benchmarking measures for EAs considering issues such as noise, computational cost, multi-objective problems?
  • Are there ways to draw conclusions on strengths and weaknesses of EA performance by increasing the interpretability of benchmark problems, e.g. by doing landscape analysis / visualisation or tying them to real-world problems?

If you are interested in this discussion, please send an email to v.volz@qmul.ac.uk. This is especially true if you are not able to attend the meeting, we will send results and updates via email.