DC Field | Value | Language |
---|---|---|
dc.contributor.author | Yoo, Jaemin | ko |
dc.contributor.author | Kim, Junghun | ko |
dc.contributor.author | Yoon, Hoyoung | ko |
dc.contributor.author | Kim, Geonsoo | ko |
dc.contributor.author | Jang, Changwon | ko |
dc.contributor.author | Kang, U | ko |
dc.date.accessioned | 2023-08-16T00:00:13Z | - |
dc.date.available | 2023-08-16T00:00:13Z | - |
dc.date.created | 2023-08-15 | - |
dc.date.created | 2023-08-15 | - |
dc.date.issued | 2021-12-10 | - |
dc.identifier.citation | 21st IEEE International Conference on Data Mining, ICDM 2021, pp.827 - 836 | - |
dc.identifier.issn | 1550-4786 | - |
dc.identifier.uri | http://hdl.handle.net/10203/311554 | - |
dc.description.abstract | How can we classify graph-structured data only with positive labels? Graph-based positive-unlabeled (PU) learning is to train a binary classifier given only the positive labels when the relationship between examples is given as a graph. The problem is of great importance for various tasks such as detecting malicious accounts in a social network, which are difficult to be modeled by supervised learning when the true negative labels are absent. Previous works for graph-based PU learning assume that the prior distribution of positive nodes is known in advance, which is not true in many real-world cases. In this work, we propose GRAB (Graph-based Risk minimization with iterAtive Belief propagation), a novel end-to-end approach for graph-based PU learning that requires no class prior. GRAB models a given graph as a Markov network and runs the marginalization and update steps iteratively. The marginalization step estimates the marginals of latent variables, while the update step trains a classifier network utilizing the computed priors in the objective function. Extensive experiments on five datasets show that GRAB achieves state-of-the-art accuracy, even compared with previous methods that are given the true prior. | - |
dc.language | English | - |
dc.publisher | Institute of Electrical and Electronics Engineers Inc. | - |
dc.title | Accurate Graph-Based PU Learning without Class Prior | - |
dc.type | Conference | - |
dc.identifier.wosid | 000780454100084 | - |
dc.identifier.scopusid | 2-s2.0-85125184239 | - |
dc.type.rims | CONF | - |
dc.citation.beginningpage | 827 | - |
dc.citation.endingpage | 836 | - |
dc.citation.publicationname | 21st IEEE International Conference on Data Mining, ICDM 2021 | - |
dc.identifier.conferencecountry | NZ | - |
dc.identifier.conferencelocation | Virtual | - |
dc.identifier.doi | 10.1109/ICDM51629.2021.00094 | - |
dc.contributor.localauthor | Yoo, Jaemin | - |
dc.contributor.nonIdAuthor | Kim, Junghun | - |
dc.contributor.nonIdAuthor | Yoon, Hoyoung | - |
dc.contributor.nonIdAuthor | Kim, Geonsoo | - |
dc.contributor.nonIdAuthor | Jang, Changwon | - |
dc.contributor.nonIdAuthor | Kang, U | - |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.