Timezone: »

Poster
Improved Sleeping Bandits with Stochastic Action Sets and Adversarial Rewards
Aadirupa Saha · Pierre Gaillard · Michal Valko

Thu Jul 16 06:00 AM -- 06:45 AM &amp; Thu Jul 16 07:00 PM -- 07:45 PM (PDT) @
In this paper, we consider the problem of sleeping bandits with stochastic action sets and adversarial rewards. In this setting, in contrast to most work in bandits, the actions may not be available at all times. For instance, some products might be out of stock in item recommendation. The best existing efficient (i.e., polynomial-time) algorithms for this problem only guarantee a $O(T^{2/3})$ upper-bound on the regret. Yet, inefficient algorithms based on EXP4 can achieve $O(\sqrt{T})$. In this paper, we provide a new computationally efficient algorithm inspired by EXP3 satisfying a regret of order $O(\sqrt{T})$ when the availabilities of each action $i \in \cA$ are independent. We then study the most general version of the problem where at each round available sets are generated from some unknown arbitrary distribution (i.e., without the independence assumption) and propose an efficient algorithm with $O(\sqrt {2^K T})$ regret guarantee. Our theoretical results are corroborated with experimental evaluations.

#### Author Information

##### Aadirupa Saha (Indian Institute of Science (IISc), Bangalore)

Bio: Aadirupa Saha is currently a visiting faculty at Toyota Technological Institute at Chicago (TTIC). She obtained her PhD from the Department of Computer Science, Indian Institute of Science, Bangalore, advised by Aditya Gopalan and Chiranjib Bhattacharyya. She spent two years at Microsoft Research New York City as a postdoctoral researcher. During her PhD, Aadirupa interned at Microsoft Research, Bangalore, Inria, Paris, and Google AI, Mountain View. Her research interests include Bandits, Reinforcement Learning, Optimization, Learning theory, Algorithms. She has organized various workshops, tutorials and also served as a reviewer in top ML conferences. Research Interests: Machine Learning Theory (specifically Online Learning, Bandits, Reinforcement Learning), Optimization, Game Theory, Algorithms. She is recently interested in exploring ML problems at the intersection of Fairness, Privacy, Game theory and Mechanism design.

##### Michal Valko (DeepMind)

Michal is a machine learning scientist in DeepMind Paris, tenured researcher at Inria, and the lecturer of the master course Graphs in Machine Learning at l'ENS Paris-Saclay. Michal is primarily interested in designing algorithms that would require as little human supervision as possible. This means 1) reducing the “intelligence” that humans need to input into the system and 2) minimizing the data that humans need to spend inspecting, classifying, or “tuning” the algorithms. That is why he is working on methods and settings that are able to deal with minimal feedback, such as deep reinforcement learning, bandit algorithms, or self-supervised learning. Michal is actively working on represenation learning and building worlds models. He is also working on deep (reinforcement) learning algorithm that have some theoretical underpinning. He has also worked on sequential algorithms with structured decisions where exploiting the structure leads to provably faster learning. He received his Ph.D. in 2011 from the University of Pittsburgh under the supervision of Miloš Hauskrecht and after was a postdoc of Rémi Munos before taking a permanent position at Inria in 2012.