aiqbench analysis
This research addresses the challenge of online active feature acquisition by framing it as a combinatorial Bandits with Knapsacks problem under global budget constraints. The authors introduce LP-Chain, an algorithm that efficiently navigates the action space to achieve improved regret bounds and competitive predictive performance compared to deep reinforcement learning baselines. This work advances the theoretical understanding of cost-sensitive classification in streaming data environments where feature observation is limited.
AFA-BANDIT: Provably Near-Optimal Online Multi-Feature Classification Under Budget Constraints
arXiv:2610.07615v1 Announce Type: new Abstract: Active Feature Acquisition (AFA) is a classification problem in which an agent decides which costly features to acquire before predicting each sample's label. Unlike batch AFA, which trains a fixed policy and classifier offline on fully observed data, online AFA updates its predictor from revealed labels as samples arrive. Existing online methods either use deep reinforcement learning (RL) without performance guarantees or maximize cost-adjusted reward rather than enforce a global budget. We formulate online AFA as a combinatorial Bandits with Knapsacks (BwK) problem that couples acquisition and prediction. Unlike prior bandit-based AFA and classical BwK, our setting has combinatorial complexity, evolving rewards, a global budget, and structured side information. We obtain an improved regret upper bound over standard BwK bounds in this framework, leveraging a cardinality-aware confidence bound and the subset update structure. To avoid an exponentially large action space, we propose \emph{LP-Chain}, a variant that searches a cost-aware chain of feature subsets with a size that grows linearly with the number of features. While the regret upper bound is specific to the combinatorial framework, \emph{LP-Chain} empirically achieves comparable predictive performance. On synthetic data, \emph{LP-Chain} outperforms HEDGE-based BwK and deep RL-based online AFA baselines and scales favorably to more features.
Abstract:Active Feature Acquisition (AFA) is a classification problem in which an agent decides which costly features to acquire before predicting each sample's label. Unlike batch AFA, which trains a fixed policy and classifier offline on fully observed data, online AFA updates its predictor from revealed labels as samples arrive. Existing online methods either use deep reinforcement learning (RL) without performance guarantees or maximize cost-adjusted reward rather than enforce a global budget. We formulate online AFA as a combinatorial Bandits with Knapsacks (BwK) problem that couples acquisition and prediction. Unlike prior bandit-based AFA and classical BwK, our setting has combinatorial complexity, evolving rewards, a global budget, and structured side information. We obtain an improved regret upper bound over standard BwK bounds in this framework, leveraging a cardinality-aware confidence bound and the subset update structure. To avoid an exponentially large action space, we propose \emph{LP-Chain}, a variant that searches a cost-aware chain of feature subsets with a size that grows linearly with the number of features. While the regret upper bound is specific to the combinatorial framework, \emph{LP-Chain} empirically achieves comparable predictive performance. On synthetic data, \emph{LP-Chain} outperforms HEDGE-based BwK and deep RL-based online AFA baselines and scales favorably to more features.
Machine Learning (cs.LG)
arXiv:2610.07615 [cs.LG]
(or arXiv:2610.07615v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2610.07615 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 294 words
Related coverage
[AINews] not much happened today
Latent Space · Sep 18, 2026
The Entangled Dimensions of Decision Theory
LessWrong · Jul 30, 2026
Digital States 2026: AI Moves From Experiment to Infrastructure
govtech.com · Sep 24, 2026
Restless Bandits with Individual Penalty Constraints: Near-Optimal Indices and Deep Reinforcement Learning
arXiv · Sep 30, 2026