FIESTA: Fast IdEntification of State-of-The-Art models using adaptive bandit algorithms.
Henry B. MossAndrew MooreDavid S. LesliePaul RaysonPublished in: CoRR (2019)
Keyphrases
- significant improvement
- probabilistic model
- optimization problems
- recently developed
- machine learning algorithms
- learning algorithm
- complex systems
- adaptive algorithms
- discrete data
- free parameters
- linear models
- learning models
- experimental data
- prior knowledge
- machine learning
- computational efficiency
- computationally efficient
- computational cost
- graph theory
- learned models
- computational complexity
- data structure
- reinforcement learning
- upper confidence bound