Fast Rates for Bandit PAC Multiclass Classification
- URL: http://arxiv.org/abs/2406.12406v1
- Date: Tue, 18 Jun 2024 08:54:04 GMT
- Title: Fast Rates for Bandit PAC Multiclass Classification
- Authors: Liad Erez, Alon Cohen, Tomer Koren, Yishay Mansour, Shay Moran,
- Abstract summary: We study multiclass PAC learning with bandit feedback, where inputs are classified into one of $K$ possible labels and feedback is limited to whether or not the predicted labels are correct.
Our main contribution is in designing a novel learning algorithm for the agnostic $(varepsilon,delta)$PAC version of the problem.
- Score: 73.17969992976501
- License: http://creativecommons.org/licenses/by/4.0/
- Abstract: We study multiclass PAC learning with bandit feedback, where inputs are classified into one of $K$ possible labels and feedback is limited to whether or not the predicted labels are correct. Our main contribution is in designing a novel learning algorithm for the agnostic $(\varepsilon,\delta)$-PAC version of the problem, with sample complexity of $O\big( (\operatorname{poly}(K) + 1 / \varepsilon^2) \log (|H| / \delta) \big)$ for any finite hypothesis class $H$. In terms of the leading dependence on $\varepsilon$, this improves upon existing bounds for the problem, that are of the form $O(K/\varepsilon^2)$. We also provide an extension of this result to general classes and establish similar sample complexity bounds in which $\log |H|$ is replaced by the Natarajan dimension. This matches the optimal rate in the full-information version of the problem and resolves an open question studied by Daniely, Sabato, Ben-David, and Shalev-Shwartz (2011) who demonstrated that the multiplicative price of bandit feedback in realizable PAC learning is $\Theta(K)$. We complement this by revealing a stark contrast with the agnostic case, where the price of bandit feedback is only $O(1)$ as $\varepsilon \to 0$. Our algorithm utilizes a stochastic optimization technique to minimize a log-barrier potential based on Frank-Wolfe updates for computing a low-variance exploration distribution over the hypotheses, and is made computationally efficient provided access to an ERM oracle over $H$.
Related papers
- Almost Minimax Optimal Best Arm Identification in Piecewise Stationary Linear Bandits [55.957560311008926]
We propose a piecewise stationary linear bandit (PSLB) model where the quality of an arm is measured by its return averaged over all contexts.
PS$varepsilon$BAI$+$ is guaranteed to identify an $varepsilon$-optimal arm with probability $ge 1-delta$ and with a minimal number of samples.
arXiv Detail & Related papers (2024-10-10T06:15:42Z) - Beyond Minimax Rates in Group Distributionally Robust Optimization via a Novel Notion of Sparsity [14.396304498754688]
We show a novel notion of sparsity that we dub $(lambda, beta)$-sparsity.
We also show an adaptive algorithm which adapts to the best $(lambda, beta)$-sparsity condition that holds.
arXiv Detail & Related papers (2024-10-01T13:45:55Z) - Efficiently Learning One-Hidden-Layer ReLU Networks via Schur
Polynomials [50.90125395570797]
We study the problem of PAC learning a linear combination of $k$ ReLU activations under the standard Gaussian distribution on $mathbbRd$ with respect to the square loss.
Our main result is an efficient algorithm for this learning task with sample and computational complexity $(dk/epsilon)O(k)$, whereepsilon>0$ is the target accuracy.
arXiv Detail & Related papers (2023-07-24T14:37:22Z) - Asymptotically Optimal Pure Exploration for Infinite-Armed Bandits [4.811176167998627]
We study pure exploration with infinitely many bandit arms generated i.i.d. from an unknown distribution.
Our goal is to efficiently select a single high quality arm whose average reward is, with probability $1-delta$, within $varepsilon$ of being among the top $eta$-fraction of arms.
arXiv Detail & Related papers (2023-06-03T04:00:47Z) - Near Sample-Optimal Reduction-based Policy Learning for Average Reward
MDP [58.13930707612128]
This work considers the sample complexity of obtaining an $varepsilon$-optimal policy in an average reward Markov Decision Process (AMDP)
We prove an upper bound of $widetilde O(H varepsilon-3 ln frac1delta)$ samples per state-action pair, where $H := sp(h*)$ is the span of bias of any optimal policy, $varepsilon$ is the accuracy and $delta$ is the failure probability.
arXiv Detail & Related papers (2022-12-01T15:57:58Z) - Instance-optimal PAC Algorithms for Contextual Bandits [20.176752818200438]
In this work, we focus on the bandit problem in the $(epsilon,delta)$-$textitPAC$ setting.
We show that no algorithm can be simultaneously minimax-optimal regret minimization and instance-dependent PAC for best-arm identification.
arXiv Detail & Related papers (2022-07-05T23:19:43Z) - Cascading Bandit under Differential Privacy [21.936577816668944]
We study emphdifferential privacy (DP) and emphlocal differential privacy (LDP) in cascading bandits.
Under DP, we propose an algorithm which guarantees $epsilon$-indistinguishability and a regret of $mathcalO(fraclog3 Tepsilon)$ for an arbitrarily small $xi$.
Under ($epsilon$,$delta$)-LDP, we relax the $K2$ dependence through the tradeoff between privacy budgetepsilon$ and error probability $
arXiv Detail & Related papers (2021-05-24T07:19:01Z) - Algorithms and SQ Lower Bounds for PAC Learning One-Hidden-Layer ReLU
Networks [48.32532049640782]
We study the problem of learning one-hidden-layer ReLU networks with $k$ hidden units on $mathbbRd$ under Gaussian marginals.
For the case of positive coefficients, we give the first-time algorithm for this learning problem for $k$ up to $tildeOOmega(sqrtlog d)$.
arXiv Detail & Related papers (2020-06-22T17:53:54Z) - Model-Free Reinforcement Learning: from Clipped Pseudo-Regret to Sample
Complexity [59.34067736545355]
Given an MDP with $S$ states, $A$ actions, the discount factor $gamma in (0,1)$, and an approximation threshold $epsilon > 0$, we provide a model-free algorithm to learn an $epsilon$-optimal policy.
For small enough $epsilon$, we show an improved algorithm with sample complexity.
arXiv Detail & Related papers (2020-06-06T13:34:41Z) - Nearly Optimal Regret for Stochastic Linear Bandits with Heavy-Tailed
Payoffs [35.988644745703645]
We analyze the linear bandits with heavy-tailed payoffs, where the payoffs admit finite $1+epsilon$ moments.
We propose two novel algorithms which enjoy a sublinear regret bound of $widetildeO(dfrac12Tfrac11+epsilon)$.
arXiv Detail & Related papers (2020-04-28T13:01:38Z)
This list is automatically generated from the titles and abstracts of the papers in this site.
This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.