Reinforcement Learning with Immediate Rewards and Linear Hypotheses

By Naoki Abe, Alan W. Biermann, Philip M. Long. 2003

130 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Management Science and Operations Research

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Management Science and Operations Research

Citations in the corpus

A contextual-bandit approach to personalized news article recommendation (2010, methodology): “Assumin g the expected payoff of an arm is linear in its features, Auer [6] d escribes theLinRel algorithm that is essentially a UCB-type approach and shows that one of its variants has a regret of Õ( √ T ), a significant improvement over earlier algorithms [1].”
Unbiased offline evaluation of contextual-bandit-based news article recommendation algorithms (2010, methodology): “…for some coeﬃcient vector w ), both LinRel [4] and LinUCB [19, 10, 11] are essentially UCB-type approaches generalized to linear payoﬀ functions, and their variants have a regret of ˜ O ( √ T ), a signiﬁcant improvement over earlier algorithms [1] as well as the more general epoch-greedy algorithm.”
Fair Algorithms for Infinite and Contextual Bandits (2016, background): “…3, we show how to design fair algorithms when at most some ﬁnite number k of individuals arrives in any round (the linear contextual bandits problem [Abe et al., 2003, Auer, 2002]), as well as when m individuals may be chosen in each round (the “multiple play" introduced and studied absent…”
A unifying framework for computational reinforcement learning theory (2009, background): “2002b; Auer 2002], associative bandits [Kaelbling 1994; Auer 2002; Abe et al. 2003; Strehl et al. 2006d], and budgeted bandits [Madani et al.”
Neural Contextual Bandits with UCB-based Exploration (2019, background): “Contextual Bandits There is a line of extensive work on linear bandits (e.g., Abe et al., 2003; Auer, 2002; Abe et al., 2003; Dani et al., 2008; Rusmevichientong & Tsitsiklis, 2010; Li et al., 2010; Chu et al., 2011; Abbasi-Yadkori et al., 2011).”
Contextual Gaussian Process Bandit Optimization (2011, background): “[9] model the expected payoff for each action as a (unknown) linear function μ(s, z) = zs θ ∗ s .”
Improved Algorithms for Linear Stochastic Bandits (2011, background): “Related problems were also studied by Abe et al. (2003), Walsh et al. (2009), Dekel et al. (2010).”
Contextual Bandits with Linear Payoff Functions (2011, methodology): “Our regret lower bound improves the earlier result Ω(T K) by Abe et al. [2003]. This new result matches the best known regret upper bounds up to logarithmic factors. (1)Even in the worst case, the computation complexity of matrix inversion is no worse than that of eigen decomposition. Recently, the ”