Using a Markov network as a surrogate fitness function in a genetic algorithm.
Alexander E. I. BrownleeOlivier Regnier-CoudertJohn A. W. McCallStewart MassiePublished in: IEEE Congress on Evolutionary Computation (2010)
Keyphrases
- fitness function
- markov networks
- genetic algorithm
- genetic programming
- graphical models
- evolutionary algorithm
- genetic algorithm ga
- evolutionary computation
- maximum likelihood
- genetic operators
- belief propagation
- first order logic
- conditional random fields
- posterior probability
- bayesian networks
- crossover operator
- crossover and mutation
- probabilistic model
- document classification
- bayesian inference
- markov random field
- hidden variables
- search space
- genetic algorithm is employed
- multi objective
- mutation operator
- evolutionary search
- simulated annealing
- neural network
- evolutionary process
- conditional probabilities
- artificial neural networks
- inductive logic programming
- fitness evaluation
- penalty function
- initial population
- population size
- information extraction
- objective function
- generative model
- particle swarm optimization