Discriminative Active Learning

Daniel Gissin, Shai Shalev-Shwartz
7/15/2019
cs.LGstat.ML

Official code: dsgissin/DiscriminativeActiveLearning (205 stars, MIT, last commit 2019-09-26).

Abstract

We propose a new batch mode active learning algorithm designed for neural networks and large query batch sizes. The method, Discriminative Active Learning (DAL), poses active learning as a binary classification task, attempting to choose examples to label in such a way as to make the labeled set and the unlabeled pool indistinguishable. Experimenting on image classification tasks, we empirically show our method to be on par with state of the art methods in medium and large query batch sizes, while being simple to implement and also extend to other domains besides classification tasks. Our experiments also show that none of the state of the art methods of today are clearly better than uncertainty sampling when the batch size is relatively large, negating some of the reported results in the recent literature.

View on arXivView PDF

Code Implementations(1)

dsgissin/DiscriminativeActiveLearningOfficial100%
MIT

Code and website for DAL (Discriminative Active Learning) - a new active learning algorithm for neural networks in the batch setting. For the blog:

20555PythonJul 3, 20187 years agoMIT
active-learningadversarial-active-learningbayesian-active-learningcore-setdeep-learning+3 more

Similar Papers

Deep Batch Active Learning by Diverse, Uncertain Gradient Lower Bounds

88% match1 repo2172019

Active Learning for Convolutional Neural Networks: A Core-Set Approach

87% match1 repo2822017

Bayesian Batch Active Learning as Sparse Subset Approximation

85% match1 repo402019

Scalable Batch Acquisition for Deep Bayesian Active Learning

83% match1 repo42023

Variational Adversarial Active Learning

83% match1 repo2312019

Active Learning for Multi-class Image Classification

83% match1 repo252025

Cite this paper

@article{gissin2019discriminative,
  title  = {Discriminative Active Learning},
  author = {Daniel Gissin and Shai Shalev-Shwartz},
  year   = {2019},
  eprint = {1907.06347},
  archivePrefix = {arXiv},
  url    = {https://arxiv.org/abs/1907.06347}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/1907.06347.svg)](https://codeofpaper.com/paper/1907.06347)

Discussion