Rates of convergence in active learning
From MaRDI portal
Publication:2429933
Abstract: We study the rates of convergence in generalization error achievable by active learning under various types of label noise. Additionally, we study the general problem of model selection for active learning with a nested hierarchy of hypothesis classes and propose an algorithm whose error rate provably converges to the best achievable error among classifiers in the hierarchy at a rate adaptive to both the complexity of the optimal classifier and the noise conditions. In particular, we state sufficient conditions for these rates to be dramatically faster than those achievable by passive learning.
Recommendations
- Statistical asymptotic theory of active learning
- Convergence rates of learning algorithms by random projection
- Probability bounds for active learning in the regression problem
- Minimax Bounds for Active Learning
- Minimax Bounds for Active Learning
- The convergence rate for a \(K\)-functional in learning theory
- Algorithmic connections between active learning and stochastic convex optimization
- Rademacher complexities and bounding the excess risk in active learning
- Efficient Algorithms for General Active Learning
Cites work
- Active Learning in the Non-realizable Case
- Agnostic active learning
- Concentration inequalities and asymptotic results for ratio type empirical processes
- Estimation of dependences based on empirical data. Transl. from the Russian by Samuel Kotz
- scientific article; zbMATH DE number 1332320 (Why is no real title available?)
- scientific article; zbMATH DE number 2034518 (Why is no real title available?)
- scientific article; zbMATH DE number 893887 (Why is no real title available?)
- Learnability and the Vapnik-Chervonenkis dimension
- Local Rademacher complexities and oracle inequalities in risk minimization. (2004 IMS Medallion Lecture). (With discussions and rejoinder)
- Margin Based Active Learning
- Minimax Bounds for Active Learning
- Neural Network Learning
- On the Uniform Convergence of Relative Frequencies of Events to Their Probabilities
- Optimal aggregation of classifiers in statistical learning.
- Probability inequalities for empirical processes and a law of the iterated logarithm
- Rates of growth and sample moduli for weighted empirical processes indexed by sets
- Risk bounds for statistical learning
- Sample moduli for set-indexed Gaussian processes
- Smooth discrimination analysis
- Toward efficient agnostic learning
- Weak convergence and empirical processes. With applications to statistics
Cited in
(25)- A theory of transfer learning with applications to active learning
- Spatially-adaptive sensing in nonparametric regression
- Surrogate losses in passive and active learning
- Fast learning rates in statistical inference through aggregation
- Probability bounds for active learning in the regression problem
- Absolute convergence and error thresholds in non-active adaptive sampling
- Refined error bounds for several learning algorithms
- Algorithmic connections between active learning and stochastic convex optimization
- Active learning using smooth relative regret approximations with applications
- Active Learning in the Non-realizable Case
- Minimax Bounds for Active Learning
- Learning noisy linear classifiers via adaptive and selective sampling
- Active nearest-neighbor learning in metric spaces
- Interactive algorithms: pool, stream and precognitive stream
- An adaptive strategy for active learning with smooth decision boundary
- Adaptive sensing performance lower bounds for sparse signal detection and support estimation
- Plug-in approach to active learning
- Active learning via perfect selective classification
- Margin Based Active Learning
- Troubleshooting image segmentation models with human-in-the-loop
- Machine learning and information theory concepts towards an AI mathematician
- Recent advances in scaling-down sampling methods in machine learning
- Attribute-efficient learning of halfspaces with malicious noise: near-optimal label complexity and noise tolerance
- Stable sample compression schemes: new applications and an optimal SVM margin bound
- Efficient Human-in-the-Loop Active Learning: A Novel Framework for Data Labeling in AI Systems
This page was built for publication: Rates of convergence in active learning
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2429933)
