TY - JOUR
T1 - Mutation Rate Matters Even When Optimizing Monotonic Functions
AU - Doerr, Benjamin
AU - Jansen, Thomas
AU - Sudholt, Dirk
AU - Winzen, Carola
AU - Zarges, Christine
PY - 2013/3
Y1 - 2013/3
N2 - Extending previous analyses on function classes like linear functions, we analyze how the simple (1+1) evolutionary algorithm optimizes pseudo-Boolean functions that are strictly monotonic. These functions have the property that whenever only 0-bits are changed to 1, then the objective value strictly increases. Contrary to what one would expect, not all of these functions are easy to optimize. The choice of the constant c in the mutation probability p(n) = c/n can make a decisive difference. We show that if c < 1, then the (1+1) EA finds the optimum of every such function in {circled dash}(n log n) iterations. For c = 1, we can still prove an upper bound of O(n
3/2). However, for c ≥ 16, we present a strictly monotonic function such that the (1+1) EA with overwhelming probability needs 2Ω(n) iterations to find the optimum. This is the first time that we observe that a constant factor change of the mutation probability changes the runtime by more than a constant factor.
AB - Extending previous analyses on function classes like linear functions, we analyze how the simple (1+1) evolutionary algorithm optimizes pseudo-Boolean functions that are strictly monotonic. These functions have the property that whenever only 0-bits are changed to 1, then the objective value strictly increases. Contrary to what one would expect, not all of these functions are easy to optimize. The choice of the constant c in the mutation probability p(n) = c/n can make a decisive difference. We show that if c < 1, then the (1+1) EA finds the optimum of every such function in {circled dash}(n log n) iterations. For c = 1, we can still prove an upper bound of O(n
3/2). However, for c ≥ 16, we present a strictly monotonic function such that the (1+1) EA with overwhelming probability needs 2Ω(n) iterations to find the optimum. This is the first time that we observe that a constant factor change of the mutation probability changes the runtime by more than a constant factor.
KW - Computational complexity
KW - Evolutionary algorithms
KW - Pseudo-Boolean optimization
KW - Runtime analysis
KW - Algorithms
KW - Mutation Rate
UR - http://www.scopus.com/inward/record.url?scp=84875042081&partnerID=8YFLogxK
U2 - 10.1162/EVCO_a_00055
DO - 10.1162/EVCO_a_00055
M3 - Article
C2 - 22035497
SN - 1063-6560
VL - 21
SP - 1
EP - 27
JO - Evolutionary Computation
JF - Evolutionary Computation
IS - 1
ER -