A New Step-Size Adaptation Rule for CMA-ES Based on the Population Midpoint Fitness.
Eryk WarchulskiJaroslaw ArabasPublished in: CEC (2021)
Keyphrases
- step size
- cma es
- candidate solutions
- evolution strategy
- population size
- evolutionary algorithm
- convergence speed
- mutation operator
- fitness landscape
- convergence rate
- cost function
- evolutionary process
- covariance matrix
- differential evolution
- global search
- genetic algorithm
- fitness function
- computationally expensive
- search space
- optimization process
- faster convergence
- natural selection
- optimization problems
- steepest descent method
- variable step size
- solution space
- genetic programming
- line search
- gradient method
- particle swarm optimization algorithm
- genetic operators
- premature convergence
- image quality
- objective function
- selection operator
- optimal solution
- upper bound
- stochastic gradient descent
- evolutionary strategy
- black box
- particle swarm optimization
- final result
- wavelet coefficients
- genetic algorithm ga
- crossover operator
- machine learning