Online learning in case of unbounded losses using follow the perturbed leader algorithm

Vladimir V. V'Yugin

Результат исследований: Вклад в журналСтатьярецензирование

1 Цитирования (Scopus)

Аннотация

In this paper the sequential prediction problem with expert advice is considered for the case where losses of experts suffered at each step cannot be bounded in advance. We present some modification of Kalai and Vempala algorithm of following the perturbed leader where weights depend on past losses of the experts. New notions of a volume and a scaled fluctuation of a game are introduced. We present a probabilistic algorithm protected from unrestrictedly large one-step losses. This algorithm has the optimal performance in the case when the scaled fluctuations of one-step losses of experts of the pool tend to zero.

Язык оригиналаАнглийский
Страницы (с-по)241-266
Число страниц26
ЖурналJournal of Machine Learning Research
Том12
СостояниеОпубликовано - янв. 2011
Опубликовано для внешнего пользованияДа

Fingerprint

Подробные сведения о темах исследования «Online learning in case of unbounded losses using follow the perturbed leader algorithm». Вместе они формируют уникальный семантический отпечаток (fingerprint).

Цитировать