paper-with-me

홈 › Papers

Beyond Softmax: A New Perspective on Gradient Bandits

2025-10-04 · Emerson Melo, David Müller arxiv

We establish a link between a class of discrete choice models and the theory of online learning and multi-armed bandits. Our contributions are: (i) sublinear regret bounds for a broad algorithmic family, encompassing Exp3 as a special case; (ii) a new class of adversarial bandit algorithms derived from generalized nested logit models \citep{wen:2001}; and (iii) \textcolor{black}{we introduce a novel class of generalized gradient bandit algorithms that extends beyond the widely used softmax formulation. By relaxing the restrictive independence assumptions inherent in softmax, our framework accommodates correlated learning dynamics across actions, thereby broadening the applicability of gradient bandit methods.} Overall, the proposed algorithms combine flexible model specification with computational efficiency via closed-form sampling probabilities. Numerical experiments in stochastic bandit settings demonstrate their practical effectiveness.

📄 PDF Abstract BibTeX arXiv:2510.03979

Code (0)

등록된 구현이 없습니다.

Tasks

Computational EfficiencyMulti-Armed Bandits

Similar Papers 제목 키워드 기반

A Lyapunov Analysis of Softmax Policy Gradient for Stochastic Bandits

2026-03-27 · Tor Lattimore arxiv

We adapt the analysis of policy gradient for continuous time $k$-armed stochastic bandits by Lattimore (2026) to the standard discrete time setup. As in continuous time, we prove that with learning rate $η= O(Δ_{\min}^2/…

An Alternate Policy Gradient Estimator for Softmax Policies

2021-12-22 · Shivam Garg, Samuele Tosatto, Yangchen Pan, Martha White 외

Policy gradient (PG) estimators are ineffective in dealing with softmax policies that are sub-optimally saturated, which refers to the situation when the policy concentrates its probability mass on sub-optimal actions. S…

Bandits with Knapsacks beyond the Worst Case

2021-12-01 · NeurIPS 2021 12 · Karthik Abinav Sankararaman, Aleksandrs Slivkins

Bandits with Knapsacks (BwK) is a general model for multi-armed bandits under supply/budget constraints. While worst-case regret bounds for BwK are well-understood, we present three results that go beyond the worst-case …

Multi-Armed Bandits

Bandits with Knapsacks beyond the Worst-Case

2020-02-01 · Karthik Abinav Sankararaman, Aleksandrs Slivkins

Bandits with Knapsacks (BwK) is a general model for multi-armed bandits under supply/budget constraints. While worst-case regret bounds for BwK are well-understood, we present three results that go beyond the worst-case …

Multi-Armed Bandits

Beyond Softmax: A Natural Parameterization for Categorical Random Variables

2025-09-29 · Alessandro Manenti, Cesare Alippi arxiv

Latent categorical variables are frequently found in deep learning architectures. They can model actions in discrete reinforcement-learning environments, represent categories in latent-variable models, or express relatio…

Graph structure learningReinforcement Learning