A Nearly Optimal Contextual Bandit Algorithm.
Mohammadreza Mohaghegh NeyshabouriKaan GökcesuSelami CiftciSuleyman Serdar KozatPublished in: CoRR (2016)
Keyphrases
- optimal solution
- dynamic programming
- learning algorithm
- cost function
- np hard
- objective function
- preprocessing
- contextual bandit
- globally optimal
- exhaustive search
- computational cost
- k means
- worst case
- detection algorithm
- probabilistic model
- information extraction
- simulated annealing
- particle swarm optimization
- clustering method
- search space
- knowledge base
- optimal path
- genetic algorithm