Investigation of training performance of convolutional neural networks evolved by genetic algorithms using an activity function.
Job Isaac BetereHiroshi KinjoKunihiko NakazonoNaoki OshiroPublished in: Artif. Life Robotics (2020)
Keyphrases
- convolutional neural networks
- genetic algorithm
- neural network
- artificial neural networks
- evolutionary algorithm
- multi objective
- training phase
- particle swarm optimization
- evolutionary process
- simulated annealing
- piecewise linear
- training samples
- stochastic search algorithms
- convolutional network
- penalty function
- training algorithm
- optimization method
- genetic programming
- supervised learning
- object detection
- fuzzy logic
- training set
- search algorithm
- feature extraction