CaltechAUTHORS
  A Caltech Library Service

How to Query An Oracle? Efficient Strategies to Label Data

Lahouti, Farshad and Kostina, Victoria and Hassibi, Babak (2021) How to Query An Oracle? Efficient Strategies to Label Data. . (Unpublished) https://resolver.caltech.edu/CaltechAUTHORS:20220804-201317566

[img] PDF - Accepted Version
See Usage Policy.

829kB

Use this Persistent URL to link to this item: https://resolver.caltech.edu/CaltechAUTHORS:20220804-201317566

Abstract

We consider the basic problem of querying an expert oracle for labeling a dataset in machine learning. This is typically an expensive and time consuming process and therefore, we seek ways to do so efficiently. The conventional approach involves comparing each sample with (the representative of) each class to find a match. In a setting with N equally likely classes, this involves N/2 pairwise comparisons (queries per sample) on average. We consider a k-ary query scheme with k ≥ 2 samples in a query that identifies (dis)similar items in the set while effectively exploiting the associated transitive relations. We present a randomized batch algorithm that operates on a round-by-round basis to label the samples and achieves a query rate of O(N/k²). In addition, we present an adaptive greedy query scheme, which achieves an average rate of ≈0.2N queries per sample with triplet queries. For the proposed algorithms, we investigate the query rate performance analytically and with simulations. Empirical studies suggest that each triplet query takes an expert at most 50% more time compared with a pairwise query, indicating the effectiveness of the proposed k-ary query schemes. We generalize the analyses to nonuniform class distributions when possible.


Item Type:Report or Paper (Discussion Paper)
Related URLs:
URLURL TypeDescription
http://arxiv.org/abs/2110.02341arXivDiscussion Paper
https://resolver.caltech.edu/CaltechAUTHORS:20221031-572094900.1Related ItemJournal Article
ORCID:
AuthorORCID
Lahouti, Farshad0000-0002-8729-873X
Kostina, Victoria0000-0002-2406-7440
Hassibi, Babak0000-0002-1375-5838
Additional Information:The authors wish to acknowledge O. Shokrollahi and the participants who helped with the experiments reported in Section VI.
Subject Keywords:Machine learning, labeling datasets, clustering, classification, entity resolution
Record Number:CaltechAUTHORS:20220804-201317566
Persistent URL:https://resolver.caltech.edu/CaltechAUTHORS:20220804-201317566
Usage Policy:No commercial reproduction, distribution, display or performance rights in this work are provided.
ID Code:116134
Collection:CaltechAUTHORS
Deposited By: George Porter
Deposited On:09 Aug 2022 17:33
Last Modified:07 Nov 2022 21:24

Repository Staff Only: item control page