activity
20242026
collaborators

9 papers

cs.LG2026

Finite and Corruption-Robust Regret Bounds in Online Inverse Linear Optimization under M-Convex Action Sets

Taihei Oki, Shinsaku Sakaue

We study online inverse linear optimization, also known as contextual recommendation, where a learner sequentially infers an agent's hidden objective vector from observed optimal a…

math.CO2026

Generalizing the Multiple Exchange Property for Matroid Bases

Taihei Oki, Tamás Schwarcz

The multiple exchange property for matroid bases states that for any bases and of a matroid and any subset , there exists a subset $Y\subseteq B\se…

cs.GT2026

Ascending Auctions for Combinatorial Markets with Frictions: A Unified Framework via Discrete Convex Analysis

Taihei Oki, Ryosuke Sato

We develop a unified ascending-auction framework for computing Walrasian equilibria in combinatorial markets with strong substitutes valuations and piecewise-linear payment functio…

cs.LG2025

No-Regret M-Concave Function Maximization: Stochastic Bandit Algorithms and Hardness of Adversarial Full-Information Setting

Taihei Oki, Shinsaku Sakaue

M-concave functions, a.k.a. gross substitute valuation functions, play a fundamental role in many fields, including discrete mathematics and economics. In practice,…

cs.LG2025

Online Inverse Linear Optimization: Efficient Logarithmic-Regret Algorithm, Robustness to Suboptimality, and Lower Bound

Shinsaku Sakaue, Taira Tsuchiya, Han Bao +1

In online inverse linear optimization, a learner observes time-varying sets of feasible actions and an agent's optimal actions, selected by solving linear optimization over the fea…

math.OC2025

Algorithmic aspects of semistability of quiver representations

Yuni Iwamasa, Taihei Oki, Tasuku Soma

We study the semistability of quiver representations from an algorithmic perspective. We present efficient algorithms for several fundamental computational problems on the semistab…