A new efficient training strategy for deep neural networks by hybridization of artificial bee colony and limited-memory BFGS optimization algorithms.
Hasan BademAlper BastürkAbdullah CaliskanMehmet Emin YükselPublished in: Neurocomputing (2017)
Keyphrases
- limited memory
- neural network
- artificial bee colony
- optimization algorithm
- data streams
- memory space
- sliding window
- influence diagrams
- real time
- quasi newton method
- optimization problems
- genetic algorithm ga
- artificial bee colony algorithm
- multi objective
- artificial neural networks
- neural network training
- special case
- convergence rate
- fuzzy logic
- evolutionary algorithm
- partially observable markov decision processes
- multiobjective optimization
- hybrid algorithm
- memory requirements
- swarm intelligence
- decision problems
- differential evolution
- computational intelligence