Hutter, MarcusPoland, Jan2015-12-101532-4435http://hdl.handle.net/1885/57847When applying aggregating strategies to Prediction with Expert Advice (PEA), the learning rate must be adaptively tuned. The natural choice of √complexity/current loss renders the analysis of Weighted Majority (WM) derivatives quite complicated. In partCopyright Information: © 2005 Marcus Hutter and Jan Poland. http://www.sherpa.ac.uk/romeo/issn/1532-4435/..."Publisher's version/PDF may be used. On open access repositories" from SHERPA/RoMEO site (as at 1/09/15).Keywords: Algorithms; Boundary value problems; Expert systems; Hierarchical systems; Learning systems; Online systems; Perturbation techniques; Probability; Adaptive adversary; Adaptive learning rate; Expected and high probability bounds; Follow the perturbed leade Adaptive adversary; Adaptive learning rate; Expected and high probability bounds; Follow the perturbed leader; General alphabet and loss; General weights; Hierarchy of experts; Online sequential prediction; Prediction with expert adviceAdaptive Online Prediction by Following the Perturbed Leader20052016-02-24