ABSTRACT
This work proposes a test function to study overlapping. The test function provides full controllability over overlapping. To achieve full controllability, the building block assigning problem is reduced to a bipartite matching problem which allow us to directly assign extent of overlapping to each gene. At the end, an experiment on overlapping shows that to four chosen crossover methods, the problem difficulty increases exponentially with the extent of overlapping.
- M. Tsuji, M. Munetomo, and K. Akama. A crossover for complex building blocks overlapping. In Proceedings of the 8th annual conference on Genetic and evolutionary computation, GECCO '06, pages 1337--1344. ACM, 2006. Google ScholarDigital Library
Index Terms
- A test function with full controllability over overlapping: estimation of distribution algorithms
Recommendations
The use of explicit building blocks in evolutionary computation
This paper proposes a new algorithm to identify and compose building blocks. Building blocks are interpreted as common subsequences between good individuals. The proposed algorithm can extract building blocks from a population explicitly. Explicit ...
Supply of Order-1 Building Blocks for Functions Linearly Combined of Sinusoidal Bases with Integral Frequencies
IJCCI 2015: Proceedings of the 7th International Joint Conference on Computational IntelligenceIn line with the theory of schema sampling, a hypothesis could be made that sufficient supply of loworder
building blocks (BBs) was one of the necessary conditions for a genetic algorithm(GA) to work.
A consequential question of this hypothesis regards, ...
Non-Exhaustive, Overlapping Co-Clustering
CIKM '17: Proceedings of the 2017 ACM on Conference on Information and Knowledge ManagementThe goal of co-clustering is to simultaneously identify a clustering of the rows as well as the columns of a two dimensional data matrix. Most existing co-clustering algorithms are designed to find pairwise disjoint and exhaustive co-clusters. However, ...
Comments