An effective algorithm for hyperparameter optimization of neural networks.
Gonzalo I. DiazAchille Fokoue-NkoutcheGiacomo NanniciniHorst SamulowitzPublished in: IBM J. Res. Dev. (2017)
Keyphrases
- neural network
- optimization algorithm
- computational complexity
- experimental evaluation
- dynamic programming
- cost function
- constrained optimization
- detection algorithm
- preprocessing
- worst case
- matching algorithm
- optimization method
- times faster
- objective function
- convergence rate
- improved algorithm
- stochastic gradient
- expectation maximization
- learning algorithm
- np hard
- similarity measure
- particle swarm optimization
- fuzzy logic
- computationally efficient
- computational cost
- ant colony optimization
- multi objective
- k means
- evolutionary algorithm
- combinatorial optimization
- pattern recognition
- optimal solution
- optimization model
- decision trees