TY - JOUR
T1 - The Unrestricted Black-Box Complexity of Jump Functions
AU - Buzdalov, Maxim
AU - Doerr, Benjamin
AU - Kever, Mikhail
N1 - Publisher Copyright:
© 2016 by the Massachusetts Institute of Technology.
PY - 2016/12/1
Y1 - 2016/12/1
N2 - We analyze the unrestricted black-box complexity of the JUMP function classes for different jump sizes. For upper bounds, we present three algorithms for small, medium, and extreme jump sizes. We prove a matrix lower bound theorem which is capable of giving better lower bounds than the classic information theory approach. Using this theorem, we prove lower bounds that almost match the upper bounds. For the case of extreme jump functions, which apart from the optimum reveal only the middle fitness value(s), we use an additional lower bound argument to show that any black-box algorithm does not gain significant insight about the problem instance from the first Ω(√en) fitness evaluations. This, together with our upper bound, shows that the black-box complexity of extreme jump functions is n + Θ(√ n).
AB - We analyze the unrestricted black-box complexity of the JUMP function classes for different jump sizes. For upper bounds, we present three algorithms for small, medium, and extreme jump sizes. We prove a matrix lower bound theorem which is capable of giving better lower bounds than the classic information theory approach. Using this theorem, we prove lower bounds that almost match the upper bounds. For the case of extreme jump functions, which apart from the optimum reveal only the middle fitness value(s), we use an additional lower bound argument to show that any black-box algorithm does not gain significant insight about the problem instance from the first Ω(√en) fitness evaluations. This, together with our upper bound, shows that the black-box complexity of extreme jump functions is n + Θ(√ n).
KW - Black-box complexity
KW - Information theory
KW - Jump functions
UR - http://www.scopus.com/inward/record.url?scp=85002045780&partnerID=8YFLogxK
U2 - 10.1162/EVCO_a_00185
DO - 10.1162/EVCO_a_00185
M3 - Article
SN - 1063-6560
VL - 24
SP - 719
EP - 744
JO - Evolutionary Computation
JF - Evolutionary Computation
IS - 4
ER -