Zobrazeno 1 - 10
of 26
pro vyhledávání: '"Tsuchiya, Taira"'
Autor:
Tsuchiya, Taira, Ito, Shinji
Follow-the-Regularized-Leader (FTRL) is a powerful framework for various online learning problems. By designing its regularizer and learning rate to be adaptive to past observations, FTRL is known to work adaptively to various properties of an underl
Externí odkaz:
http://arxiv.org/abs/2405.20028
Follow-The-Regularized-Leader (FTRL) is known as an effective and versatile approach in online learning, where appropriate choice of the learning rate is crucial for smaller regret. To this end, we formulate the problem of adjusting FTRL's learning r
Externí odkaz:
http://arxiv.org/abs/2403.00715
Autor:
Tsuchiya, Taira, Ito, Shinji
In this paper, we explore online convex optimization (OCO) and introduce a new analysis that provides fast rates by exploiting the curvature of feasible sets. In online linear optimization, it is known that if the average gradient of loss functions i
Externí odkaz:
http://arxiv.org/abs/2402.12868
Autor:
Ito, Kaito, Tsuchiya, Taira
This paper investigates the problem of controlling a linear system under possibly unbounded and degenerate noise with unknown cost functions, known as an online control problem. In contrast to the existing work, which assumes the boundedness of noise
Externí odkaz:
http://arxiv.org/abs/2402.10252
Partial monitoring is a generic framework of online decision-making problems with limited observations. To make decisions from such limited observations, it is necessary to find an appropriate distribution for exploration. Recently, a powerful approa
Externí odkaz:
http://arxiv.org/abs/2402.08321
This paper studies online structured prediction with full-information feedback. For online multiclass classification, Van der Hoeven (2020) established \emph{finite} surrogate regret bounds, which are independent of the time horizon, by introducing a
Externí odkaz:
http://arxiv.org/abs/2402.08180
Autor:
Tsuchiya, Taira
甲第24939号
情博第850号
新制||情||142(附属図書館)
学位規則第4条第1項該当
Doctor of Informatics
Kyoto University
DFAM
情博第850号
新制||情||142(附属図書館)
学位規則第4条第1項該当
Doctor of Informatics
Kyoto University
DFAM
Externí odkaz:
http://hdl.handle.net/2433/285873
We study best-of-both-worlds algorithms for $K$-armed linear contextual bandits. Our algorithms deliver near-optimal regret bounds in both the adversarial and stochastic regimes, without prior knowledge about the environment. In the stochastic regime
Externí odkaz:
http://arxiv.org/abs/2312.15433
Adaptivity to the difficulties of a problem is a key property in sequential decision-making problems to broaden the applicability of algorithms. Follow-the-regularized-leader (FTRL) has recently emerged as one of the most promising approaches for obt
Externí odkaz:
http://arxiv.org/abs/2305.17301
This study considers the partial monitoring problem with $k$-actions and $d$-outcomes and provides the first best-of-both-worlds algorithms, whose regrets are favorably bounded both in the stochastic and adversarial regimes. In particular, we show th
Externí odkaz:
http://arxiv.org/abs/2207.14550