Adaptively learning the crowd kernel
Research output: Contribution to journal › Conference article › Research › peer-review
Standard
Adaptively learning the crowd kernel. / Tamuz, Omer; Liu, Ce; Belongie, Serge; Shamir, Ohad; Kalai, Adam Tauman.
In: Proceedings of the 28th International Conference on Machine Learning, ICML 2011, 2011, p. 673-680.Research output: Contribution to journal › Conference article › Research › peer-review
Harvard
APA
Vancouver
Author
Bibtex
}
RIS
TY - GEN
T1 - Adaptively learning the crowd kernel
AU - Tamuz, Omer
AU - Liu, Ce
AU - Belongie, Serge
AU - Shamir, Ohad
AU - Kalai, Adam Tauman
PY - 2011
Y1 - 2011
N2 - We introduce an algorithm that, given n objects, learns a similarity matrix over all n2 pairs, from crowdsourced data alone. The algorithm samples responses to adaptively chosen triplet-based relative-similarity queries. Each query has the form "is object a more similar to b or to c?" and is chosen to be maximally informative given the preceding responses. The output is an embedding of the objects into Euclidean space (like MDS); we refer to this as the "crowd kernel." SVMs reveal that the crowd kernel captures prominent and subtle features across a number of domains, such as "is striped" among neckties and "vowel vs. consonant" among letters.
AB - We introduce an algorithm that, given n objects, learns a similarity matrix over all n2 pairs, from crowdsourced data alone. The algorithm samples responses to adaptively chosen triplet-based relative-similarity queries. Each query has the form "is object a more similar to b or to c?" and is chosen to be maximally informative given the preceding responses. The output is an embedding of the objects into Euclidean space (like MDS); we refer to this as the "crowd kernel." SVMs reveal that the crowd kernel captures prominent and subtle features across a number of domains, such as "is striped" among neckties and "vowel vs. consonant" among letters.
UR - http://www.scopus.com/inward/record.url?scp=80053456767&partnerID=8YFLogxK
M3 - Conference article
AN - SCOPUS:80053456767
SP - 673
EP - 680
JO - Proceedings of the 28th International Conference on Machine Learning, ICML 2011
JF - Proceedings of the 28th International Conference on Machine Learning, ICML 2011
T2 - 28th International Conference on Machine Learning, ICML 2011
Y2 - 28 June 2011 through 2 July 2011
ER -
ID: 301831017