Free Lunch for Optimisation under the Universal Distribution
2014 (English)In: 2014 IEEE Congress on Evolutionary Computation (CEC), New York: IEEE Computer Society, 2014, 167-174 p.Conference paper (Refereed)
Function optimisation is a major challenge in computer science. The No Free Lunch theorems state that if all functions with the same histogram are assumed to be equally probable then no algorithm outperforms any other in expectation. We argue against the uniform assumption and suggest a universal prior exists for which there is a free lunch, but where no particular class of functions is favoured over another. We also prove upper and lower boundson the size of the free lunch.
Place, publisher, year, edition, pages
New York: IEEE Computer Society, 2014. 167-174 p.
No free Lunch, Black-box optimisation, universal distribution, Solomonoff induction, Kolmogorov complexity
Computer Science Algebra and Logic Robotics
Research subject Computer Science; Mathematics
IdentifiersURN: urn:nbn:se:su:diva-112781ISI: 000356684600024ISBN: 978-1-4799-1488-3OAI: oai:DiVA.org:su-112781DiVA: diva2:780777
IEEE Congress on Evolutionary Computation (CEC), July 6-11, 2014, Beijing, People's Republic of China