How to use expert advice

  title={How to use expert advice},
  author={N. Cesa-Bianchi and Y. Freund and D. Haussler and D. Helmbold and R. Schapire and Manfred K. Warmuth},
  journal={J. ACM},
  • N. Cesa-Bianchi, Y. Freund, +3 authors Manfred K. Warmuth
  • Published 1997
  • Mathematics, Computer Science
  • J. ACM
  • We analyze algorithms that predict a binary value by combining the predictions of several prediction strategies, called experts. Our analysis is for worst-case situations, i.e., we make no assumptions about the way the sequence of bits to be predicted is generated. We measure the performance of the algorithm by the difference between the expected number of mistakes it makes on the bit sequence and the expected number of mistakes made by the best expert on this sequence, where the expectation is… CONTINUE READING
    208 Citations

    Figures and Topics from this paper

    Lipschitz Adaptivity with Multiple Learning Rates in Online Learning
    • 7
    • Highly Influenced
    • PDF
    Tight Lower Bounds for the Multiplicative Weights Algorithm on the Experts Problem
    • Highly Influenced
    • PDF
    Non-asymptotic calibration and resolution
    • 17
    • Highly Influenced
    • PDF
    On delayed prediction of individual sequences
    • 31
    • Highly Influenced
    • PDF
    Genral Linear Relations among Different Types of Predictive Complexity
    • 13
    • Highly Influenced
    The robustness of the p-norm algorithms
    • 139
    • Highly Influenced
    Structured Weight-Based Prediction Algorithms
    • Highly Influenced
    Finite-time 4-expert prediction problem
    • 5
    • Highly Influenced
    • PDF
    Online Submodular Maximization under a Matroid Constraint with Application to Learning Assignments
    • 24
    • Highly Influenced
    • PDF
    Good Listeners, Wise Crowds, and Parasitic Experts
    • Highly Influenced
    • PDF


    Aggregating strategies
    • V. Vovk
    • Computer Science
    • COLT '90
    • 1990
    • 693
    • Highly Influential