N. Cesa-Bianchi, Y. Freund, D. P. Helmbold, and M. Warmuth. On-line prediction and conversion strategies. In Computational Learning Theory: Eurocolt '93, pages 205#216, Oxford, 1994. Oxford University Press. Home/Search Document Details and Download
Summary Related Articles Check |
This paper is cited in the following contexts:
On Optimal Sequential Prediction for General Processes - Nobel (2001) (Correct)
....decisions of the individual schemes in F . Aggregating methods, and corresponding bounds on the di erence between the loss of the aggregate scheme and that of the best scheme in the family, have been established in a variety of settings. Representative work and further references can be found in [41, 17, 27, 10, 9, 25]. Foster and Vohra [19] give an account of the aggregating problem and its history. Merhav and Feder [28] give an overview of prediction from individual sequences. Weissman and Merhav [43] establish nite sample aggregation bounds for the prediction of individual binary sequences observed in ....
N. Cesa-Bianchi, Y. Freund, D.P. Helmbold, and M.K. Warmuth, On-line prediction and conversion strategies, Machine Learning, vol.25, pp.71-110, 1996.
On Consulting a Set of Experts and Searching - Galperin (1996) (Correct)
....under a variety of assumptions. The on line problem is distinguished by the manager having to reach a decision in every round, before seeing all inputs. Littlestone and Warmuth s [44] WM (Weighted Majority) algorithm and later the BW (Binomial Weighting) algorithm by Cesa Bianchi et al. [18] address the question of consulting a finite number of experts that provide boolean advice. The prediction domain of either the manager or the experts may be modified to be the real interval [0; 1] as in Cesa Bianchi et al. 17] Littlestone and Warmuth [44] Haussler et al. 32] and Vovk [70] ....
....in every round of the game, rather than only in rounds in which the manager errs. We prove the PM algorithms achieve similar bounds when the decision domain is a finite set of arbitrary size. This generalizes the problem previously addressed by Littlestone and Warmuth [44] and Cesa Bianchi et al. [18] of decision domains of size two ( yes no questions) The PM algorithms can also be applied to tracing good sets of experts of arbitrary size 1, if such are known to exist, and allow the manager to incorporate a non uniform prior on experts quality. Expert Games explored herein are closely ....
[Article contains additional citation context not shown here]
N. Cesa-Bianchi, Y. Freund, D.P. Helmbold, and M. Warmuth. On-line prediction and conversion strategies. In EuroCOLT, pages 205--216. Clarendon Press, 1993.
Continuous Drifting Games - Freund, Opper (2000) Self-citation (Freund) (Correct)
....of the sheep are s i , the shepherd suffers the final average loss: L = 2. 2 Why are drifting games interesting Drifting games were introduced by Schapire in [13] as an abstraction which generalizes the boost by majority algorithm [7] and the binomial weights learning algorithm [5]. In this subsection we describe these relations and provide the motivation for studying drifting games. In the next subsection we motivate the study of continuous drifting games. The boost by majority algorithm corresponds to a very simple drifting game, the adversary controls the weak learning ....
....Interestingly, this very same game, with a different interpretation for sheep and locations, corresponds to the problem of online learning with expert advice as well as to an interesting variant of the twenty one questions game. The online learning problem was studied by Cesa Binachi et. al in [5]. In this case the sheep correspond to the experts and the location of the sheep corresponds to the number of mistakes made by the experts. An assumption is made that there is an expert which makes no more than k mistakes, the identity of this expert is unknown. The goal is to design an algorithm ....
Nicolo Cesa-Bianchi, Yoav Freund, David P. Helmbold, and Manfred K. Warmuth. On-line prediction and conversion strategies. Machine Learning, 25:71--110, 1996.
Merl A Mitsubishi Electric Research Laboratory - Http Www Merl (1996) (Correct)
No context found.
N. Cesa-Bianchi, Y. Freund, D. P. Helmbold, and M. Warmuth. On-line prediction and conversion strategies. In Computational Learning Theory: Eurocolt '93, pages 205#216, Oxford, 1994. Oxford University Press.
Universal Prediction - Merhav, Feder (1998) (17 citations) (Correct)
No context found.
N. Cesa-Bianchi, Y. Freund, D.P. Helmbold, and M.K. Warmuth, "On-line prediction and conversion strategies," Proc. EUROCOLT '93, pp. 205--216, Oxford, 1993.
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC