[2006.13999] Minimum Cost Active Labelingopen searchopen navigation menucontact arXivsubscribe to arXiv mailings

Labeling a data set completely is important for groundtruth generation. In this paper, we consider the problem of minimum-cost labeling: classifying all images in a large data set with a target accuracy bound at minimum dollar cost. Human labeling can be prohibitive, so we train a classifier to accurately label part of the data set. However, training the classifier can be expensive too, particularly with active learning. Our min-cost labeling uses a variant of active learning to learn a model to predict the optimal training set size for the classifier that minimizes overall cost, then uses active learning to train the classifier to maximize the number of samples the classifier can correctly label. We validate our approach on well-known public data sets such as Fashion, CIFAR-10, and CIFAR-100. In some cases, our approach has 6X lower overall cost relative to human labeling, and is always cheaper than the cheapest active learning strategy.

2 mentions: @jbohnslav
Date: 2020/06/27 08:21

Referring Tweets

@jbohnslav Minimum cost active learning arxiv: t.co/5amGQqQ1Hj Jointly optimize for minimizing human labeling cost and GPU cost. seems super useful! t.co/WU7CdVwwVP

Related Entries

Read more Social Implementation of Machine Learning - Speaker Deck
24 users, 19 mentions 2020/07/10 09:51
Read more 機械学習を「社会実装」するということ / Social Implementation of Machine Learning - Speaker Deck
0 users, 67 mentions 2020/07/11 18:52
Read more 世界初!教師データなしで高次元データの特徴を正確に獲得できるAI技術を開発 : 富士通
5 users, 28 mentions 2020/07/13 03:51
Read more AIにおける「次元の呪い」解決へ、富士通研が機械学習の最有力学会で発表 | 日経クロステック(xTECH)
2 users, 15 mentions 2020/07/13 06:52
Read more AIにおける「次元の呪い」解決へ、富士通研が機械学習の最有力学会で発表 | 日経クロステック(xTECH)
0 users, 103 mentions 2020/07/13 09:51