Computational limitations on learning from examples
Journal of the ACM (JACM)
Noise strategies for improving local search
AAAI '94 Proceedings of the twelfth national conference on Artificial intelligence (vol. 1)
Experimental results on the crossover point in random 3-SAT
Artificial Intelligence - Special volume on frontiers in problem solving: phase transitions and complexity
A solvable model of a hard optimisation problem
Theoretical aspects of evolutionary computing
Phase Transitions and Backbones of 3-SAT and Maximum 3-SAT
CP '01 Proceedings of the 7th International Conference on Principles and Practice of Constraint Programming
On the Analysis of Evolutionary Algorithms - A Proof That Crossover Really Can Help
ESA '99 Proceedings of the 7th Annual European Symposium on Algorithms
Finding critical backbone structures with genetic algorithms
Proceedings of the 9th annual conference on Genetic and evolutionary computation
A building-block royal road where crossover is provably essential
Proceedings of the 9th annual conference on Genetic and evolutionary computation
Complexity of Max-SAT using stochastic algorithms
Proceedings of the 10th annual conference on Genetic and evolutionary computation
Symmetry breaking in population-based optimization
IEEE Transactions on Evolutionary Computation
Large Barrier Trees for Studying Search
IEEE Transactions on Evolutionary Computation
Improving Performance in Combinatorial Optimisation Using Averaging and Clustering
EvoCOP '09 Proceedings of the 9th European Conference on Evolutionary Computation in Combinatorial Optimization
Benefits of a population: five mechanisms that advantage population-based algorithms
IEEE Transactions on Evolutionary Computation
Hyperplane initialized local search for MAXSAT
Proceedings of the 15th annual conference on Genetic and evolutionary computation
On handling ephemeral resource constraints in evolutionary search
Evolutionary Computation
Hi-index | 0.00 |
A new algorithm for solving maximum satisfiability (MAX-SAT) problems is introduced which clusters good solutions, and restarts the search from the closest feasible solution to the centroid of each cluster. This is shown to be highly efficient for finding good solutions of large MAX-SAT problems. We argue that this success is due to the population learning the large-scale strncture of the fitness landscape. Systematic studies of the landscape are presented to support this hypothesis. In addition, a number of other strategies are tested to rule out other possible explanations of the success. Preliminary results are shown, indicating that extensions of the proposed algorithm can give similar improvements on other hard optimization problems.