Active Online Learning with Hidden Shifting Domains

Abstract

Online machine learning systems need to adapt to domain shifts. Meanwhile, acquiring label at every timestep is expensive. We propose a surprisingly simple algorithm that adaptively balances its regret and its number of label queries in settings where the data streams are from a mixture of hidden domains. For online linear regression with oblivious adversaries, we provide a tight tradeoff that depends on the durations and dimensionalities of the hidden domains. Our algorithm can adaptively deal with interleaving spans of inputs from different domains. We also generalize our results to non-linear regression for hypothesis classes with bounded eluder dimension and adaptive adversaries. Experiments on synthetic and realistic datasets demonstrate that our algorithm achieves lower regret than uniform queries and greedy queries with equal labeling budget.

References in corpus (9)

• Do ImageNet Classifiers Generalize to ImageNet?
• Do CIFAR-10 Classifiers Generalize to CIFAR-10?
• Finite-Time Analysis of Kernelised Contextual Bandits
• Agnostic Active Learning Without Constraints
• Leveraged volume sampling for linear regression
• Active Regression by Stratification
• The Extended Littlestone's Dimension for Learning with Mistakes and Abstentions
• Strongly Adaptive Online Learning
• Adaptivity to Noise Parameters in Nonparametric Active Learning