Untitled

Vovk, Vladimir

doi:10.1023/a:1007595032382

Cited by 48 publications

(9 citation statements)

References 16 publications

(37 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The proof of this theorem is given in Section 5. Note that while the algorithm may seem to be designed for the stochastic setting, we apply it to the pure adversarial case 9 and obtain the loss guarantees. At the same time, the adversarial loss bound (5) depends on the probability distribution p(•) for which the algorithm is designed.…”

Section: Guarantees Of Performancementioning

confidence: 99%

See 1 more Smart Citation

Adaptive hedging under delayed feedback

2020

View full text Add to dashboard Cite

The article is devoted to investigating the application of hedging strategies to online expert weight allocation under delayed feedback. As the main result we develop the General Hedging algorithm G based on the exponential reweighing of experts' losses. We build the artificial probabilistic framework and use it to prove the adversarial loss bounds for the algorithm G in the delayed feedback setting. The designed algorithm G can be applied to both countable and continuous sets of experts. We also show how algorithm G extends classical Hedge (Multiplicative Weights) and adaptive Fixed Share algorithms to the delayed feedback and derive their regret bounds for the delayed setting by using our main result.

show abstract

Section: Guarantees Of Performancementioning

confidence: 99%

“…We consider the Decision-Theoretic Online Learning (DTOL) framework [1,2,3,4,5,3] which is closely related to the paradigm of prediction with expert advice [6,7,8,9,1,10,11,12]. A master algorithm at every step t = 1, .…”

Section: Introductionmentioning

confidence: 99%

Adaptive hedging under delayed feedback

2020

View full text Add to dashboard Cite

show abstract

“…In this section we discuss basic aggregating algorithms for 1-step-ahead forecasting based on exponential reweighing. Our framework is built on the general aggregating algorithm G 1 by Vovk (1999), we discuss it in Subsection 3.1. The simplest and earliest version V 1 by Vovk (1998) of this algorithm is discussed in Subsection 3.2.…”

Section: Aggregating Algorithm For 1-step-ahead Forecastingmentioning

confidence: 99%

“…In this work, we investigate the problem of modifying aggregating algorithms based on exponential reweighing for the long-term forecasting. We consider the general aggregating algorithm by Vovk (1999) for the 1-step-ahead forecasting and provide its reasonable nonreplicated generalization for the D-th-step-ahead forecasting. These algorithms are denoted by G 1 and G D respectively.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Aggregating Strategies for Long-term Forecasting

Korotin,

V'yugin,

Burnaev

2018

Preprint

View full text Add to dashboard Cite

The article is devoted to investigating the application of aggregating algorithms to the problem of the long-term forecasting. We examine the classic aggregating algorithms based on the exponential reweighing. For the general Vovk's aggregating algorithm we provide its generalization for the long-term forecasting. For the special basic case of Vovk's algorithm we provide its two modifications for the long-term forecasting. The first one is theoretically close to an optimal algorithm and is based on replication of independent copies. It provides the time-independent regret bound with respect to the best expert in the pool. The second one is not optimal but is more practical and has O( √ T ) regret bound, where T is the length of the game.

show abstract

Adaptive and Self-Confident On-Line Learning Algorithms

Auer

Cesa-Bianchi

Gentile

2002

Journal of Computer and System Sciences

154

204

View full text Add to dashboard Cite

We study on-line learning in the linear regression framework. Most of the performance bounds for on-line algorithms in this framework assume a constant learning rate. To achieve these bounds the learning rate must be optimized based on a posteriori information. This information depends on the whole sequence of examples and thus it is not available to any strictly on-line algorithm. We introduce new techniques for adaptively tuning the learning rate as the data sequence is progressively revealed. Our techniques allow us to prove essentially the same bounds as if we knew the optimal learning rate in advance. Moreover, such techniques apply to a wide class of on-line algorithms, including p-norm algorithms for generalized linear regression and Weighted Majority for linear regression with absolute loss. Our adaptive tunings are radically different from previous techniques, such as the so-called doubling trick. Whereas the doubling trick restarts the on-line algorithm several times using a constant learning rate for each run, our methods save information by changing the value of the learning rate very smoothly. In fact, for Weighted Majority over a finite set of experts our analysis provides a better leading constant than the doubling trick. © 2002 Elsevier Science (USA)

show abstract

Untitled

Cited by 48 publications

References 16 publications

Adaptive hedging under delayed feedback

Adaptive hedging under delayed feedback

Aggregating Strategies for Long-term Forecasting

Adaptive and Self-Confident On-Line Learning Algorithms

Contact Info

Product

Resources

About