A single algorithm for both restless and rested rotting bandits.
Julien SeznecPierre MénardAlessandro LazaricMichal ValkoPublished in: AISTATS (2020)
Keyphrases
- learning algorithm
- times faster
- experimental evaluation
- cost function
- detection algorithm
- computational complexity
- expectation maximization
- high accuracy
- dynamic programming
- improved algorithm
- selection algorithm
- computational cost
- k means
- preprocessing
- neural network
- significant improvement
- search space
- particle swarm optimization
- theoretical analysis
- optimal solution
- mathematical model
- matching algorithm
- feature selection
- recognition algorithm
- np hard
- input data
- data sets