On the Complexity of Best-Arm Identification in Multi-Armed Bandit Models

By Emilie Kaufmann, Olivier Capp'e, Aurélien Garivier. 2014

1,103 citations recorded by the source corpus.

The stochastic multi-armed bandit model is a simple abstraction that has proven useful in many different contexts in statistics and machine learning. Whereas the achievable limit in terms of regret minimization is now well known, our aim is to contribute to a better understanding of the performance in terms of identifying the m best arms. We introduce generic notions of complexity for the two dominant frameworks considered in the literature: fixed-budget and fixed-confidence settings. In the fixed-confidence setting, we provide the first known distribution-dependent lower bound on the complexity that involves information-theoretic quantities and holds when m ≥ 1 under general assumptions. In the specific case of two armed-bandits, we derive refined lower bounds in both the fixedcon fidence and fixed-budget settings, along with matching algorithms for Gaussian and Bernoulli bandit models. These results show in particular that the complexity of the fixed-budget setting may be smaller than the complexity of the fixed-confidence setting, contradicting the familiar behavior observed when testing fully specified alternatives. In addition, we also provide improved sequential stopping rules that have guaranteed error probabilities and shorter average running times. The proofs rely on two technical results that are of independent interest: a deviation lemma for self-normalized sums (Lemma 7) and a novel change of measure inequality for bandit models (Lemma 1).

Subfields: Artificial Intelligence, Management Science and Operations Research, Statistical Learning Theory, Algorithms & Complexity

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

  1. Frank Hutter
  2. Michael I. Jordan
  3. Aadirupa Saha
  4. Ameet Talwalkar
  5. Tor Lattimore
  6. Sergey Levine
  7. Quoc V. Le
  8. Csaba Szepesvari

In Management Science and Operations Research

  1. Aadirupa Saha
  2. Tor Lattimore
  3. Frank Hutter
  4. Chi Jin
  5. Csaba Szepesvari
  6. Akshay Krishnamurthy
  7. Emilie Kaufmann
  8. Ameet Talwalkar

In Statistical Learning Theory

  1. Tor Lattimore
  2. Aaditya Ramdas
  3. Steven R. Howard
  4. Chi Jin
  5. Akshay Krishnamurthy
  6. Aadirupa Saha
  7. Emilie Kaufmann
  8. Aurélien Garivier

In Algorithms & Complexity

  1. Aaditya Ramdas
  2. Steven R. Howard
  3. Aadirupa Saha
  4. Emilie Kaufmann
  5. Aurélien Garivier
  6. Ameet Talwalkar
  7. Mark Harman
  8. A. Dimakis

Citations in the corpus