Are Two (Samples) Really Better Than One? On the Non-Asymptotic Performance of Empirical Revenue Maximization.
Moshe BabaioffYannai A. GonczarowskiYishay MansourShay MoranPublished in: CoRR (2018)
Keyphrases
- asymptotic properties
- training samples
- data sets
- finite sample
- machine learning
- data samples
- theoretical analysis
- asymptotically optimal
- small sample
- training set
- special case
- empirical data
- real time
- genetic algorithm
- revenue management
- market share
- dynamic pricing
- information theoretic
- sample size
- worst case
- objective function