Activized Learning: Transforming Passive to Active with Improved Label Complexity

MaRDI QIDQ5405167zbMATH OpenFDO

Authors Steve Hanneke

Publication date 1 April 2014

Full work available at URL https://arxiv.org/abs/1108.1766, http://www.jmlr.org/papers/v13/hanneke12a.html

zbMATH Keywords

active learningPAC learningsample complexityselective samplingsequential designstatistical learning theory

Mathematics Subject Classification ID

Classification and discrimination; cluster analysis (statistical aspects) (62H30) Learning and adaptive systems in artificial intelligence (68T05)

Abstract: We study the theoretical advantages of active learning over passive learning. Specifically, we prove that, in noise-free classifier learning for VC classes, any passive learning algorithm can be transformed into an active learning algorithm with asymptotically strictly superior label complexity for all nontrivial target functions and distributions. We further provide a general characterization of the magnitudes of these improvements in terms of a novel generalization of the disagreement coefficient. We also extend these results to active learning in the presence of label noise, and find that even under broad classes of noise distributions, we can typically guarantee strict improvements over the known results for passive learning.

Recommendations

• Active learning for hierarchical multi-label classification
• An approach to active learning for classifier systems
• Active learning in neural networks
• Learning cost-sensitive active classifiers
• Active learning with a misspecified prior

Cited in

(6)

• Surrogate losses in passive and active learning
• Minimax analysis of active learning
• Learning with a drifting target concept
• The relationship between agnostic selective classification, active learning and the disagreement coefficient
• Recent advances in scaling-down sampling methods in machine learning
• Stable sample compression schemes: new applications and an optimal SVM margin bound

This page was built for publication: Activized Learning: Transforming Passive to Active with Improved Label Complexity

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5405167)