RAL - Improving Stream-Based Active Learning by Reinforcement Learning

Abstract : One of the main challenges associated with supervised learning under dynamic scenarios is that of periodically getting access to labels of fresh, previously unseen samples. Labeling new data is usually an expensive and cumbersome process, while not all data points are equally valuable. Active learning aims at labeling only the most informative samples to reduce cost. In this paradigm, a learner can choose from which new samples it wants to learn, and can obtain the ground truth by asking an oracle for the corresponding labels. We introduce RAL - Reinforced stream-based Active Learning -, a new active-learning approach, coupling stream-based active learning with reinforcement-learning concepts. In particular, we model active learning as a contextual-bandit problem, in which rewards are based on the usefulness of the system's querying behavior. Empirical evaluations on multiple datasets confirm that RAL outperforms the state of the art, both by improving learning accuracy and by reducing the number of requested labels. As an additional contribution, we release RAL as an open-source Python package to the machine-learning community.
Complete list of metadatas

Cited literature [24 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-02265426
Contributor : Sarah Wassermann <>
Submitted on : Friday, August 9, 2019 - 4:11:46 PM
Last modification on : Wednesday, August 14, 2019 - 1:11:18 AM

File

ecmlpkdd2019.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-02265426, version 1

Citation

Sarah Wassermann, Thibaut Cuvelier, Pedro Casas. RAL - Improving Stream-Based Active Learning by Reinforcement Learning. European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML-PKDD) Workshop on Interactive Adaptive Learning (IAL), Sep 2019, Würzburg, Germany. ⟨hal-02265426⟩

Share

Metrics

Record views

2

Files downloads

4