A Proposal for Supervised Clustering with Dirichlet Process Using Labels. Peralta, B., Caro, A., & Soto, A. Pattern Recognition Letters, 80:52-57, 2016.
Paper abstract bibtex Supervised clustering is an emerging area of machine learning, where the goal is to find class-uniform clusters. However, typical state-of-the-art algorithms use a fixed number of clusters. In this work, we propose a variation of a non-parametric Bayesian modeling for supervised clustering. Our approach consists of modeling the clusters as a mixture of Gaussians with the constraint of encouraging clusters of points with the same label. In order to estimate the number of clusters, we assume a-priori a countably infinite number of clusters using a variation of Dirichlet Process model over the prior distribution. In our experiments, we show that our technique typically outperforms the results of other clustering techniques.
@Article{ peralta:etal:2016,
author = {B. Peralta and A. Caro and A. Soto},
title = {A Proposal for Supervised Clustering with Dirichlet
Process Using Labels},
journal = {Pattern Recognition Letters},
volume = {80},
pages = {52-57},
year = {2016},
abstract = {Supervised clustering is an emerging area of machine
learning, where the goal is to find class-uniform clusters.
However, typical state-of-the-art algorithms use a fixed
number of clusters. In this work, we propose a variation of
a non-parametric Bayesian modeling for supervised
clustering. Our approach consists of modeling the clusters
as a mixture of Gaussians with the constraint of
encouraging clusters of points with the same label. In
order to estimate the number of clusters, we assume
a-priori a countably infinite number of clusters using a
variation of Dirichlet Process model over the prior
distribution. In our experiments, we show that our
technique typically outperforms the results of other
clustering techniques.},
url = {http://www.sciencedirect.com/science/article/pii/S0167865516300976}
}
Downloads: 0
{"_id":"W7QwBa57SRTwQ7dsp","bibbaseid":"peralta-caro-soto-aproposalforsupervisedclusteringwithdirichletprocessusinglabels-2016","downloads":0,"creationDate":"2017-01-16T02:50:32.888Z","title":"A Proposal for Supervised Clustering with Dirichlet Process Using Labels","author_short":["Peralta, B.","Caro, A.","Soto, A."],"year":2016,"bibtype":"article","biburl":"https://raw.githubusercontent.com/ialab-puc/ialab.ing.puc.cl/master/pubs.bib","bibdata":{"bibtype":"article","type":"article","author":[{"firstnames":["B."],"propositions":[],"lastnames":["Peralta"],"suffixes":[]},{"firstnames":["A."],"propositions":[],"lastnames":["Caro"],"suffixes":[]},{"firstnames":["A."],"propositions":[],"lastnames":["Soto"],"suffixes":[]}],"title":"A Proposal for Supervised Clustering with Dirichlet Process Using Labels","journal":"Pattern Recognition Letters","volume":"80","pages":"52-57","year":"2016","abstract":"Supervised clustering is an emerging area of machine learning, where the goal is to find class-uniform clusters. However, typical state-of-the-art algorithms use a fixed number of clusters. In this work, we propose a variation of a non-parametric Bayesian modeling for supervised clustering. Our approach consists of modeling the clusters as a mixture of Gaussians with the constraint of encouraging clusters of points with the same label. In order to estimate the number of clusters, we assume a-priori a countably infinite number of clusters using a variation of Dirichlet Process model over the prior distribution. In our experiments, we show that our technique typically outperforms the results of other clustering techniques.","url":"http://www.sciencedirect.com/science/article/pii/S0167865516300976","bibtex":"@Article{\t peralta:etal:2016,\n author\t= {B. Peralta and A. Caro and A. Soto},\n title\t\t= {A Proposal for Supervised Clustering with Dirichlet\n\t\t Process Using Labels},\n journal\t= {Pattern Recognition Letters},\n volume\t= {80},\n pages\t\t= {52-57},\n year\t\t= {2016},\n abstract\t= {Supervised clustering is an emerging area of machine\n\t\t learning, where the goal is to find class-uniform clusters.\n\t\t However, typical state-of-the-art algorithms use a fixed\n\t\t number of clusters. In this work, we propose a variation of\n\t\t a non-parametric Bayesian modeling for supervised\n\t\t clustering. Our approach consists of modeling the clusters\n\t\t as a mixture of Gaussians with the constraint of\n\t\t encouraging clusters of points with the same label. In\n\t\t order to estimate the number of clusters, we assume\n\t\t a-priori a countably infinite number of clusters using a\n\t\t variation of Dirichlet Process model over the prior\n\t\t distribution. In our experiments, we show that our\n\t\t technique typically outperforms the results of other\n\t\t clustering techniques.},\n url\t\t= {http://www.sciencedirect.com/science/article/pii/S0167865516300976}\n}\n\n","author_short":["Peralta, B.","Caro, A.","Soto, A."],"key":"peralta:etal:2016","id":"peralta:etal:2016","bibbaseid":"peralta-caro-soto-aproposalforsupervisedclusteringwithdirichletprocessusinglabels-2016","role":"author","urls":{"Paper":"http://www.sciencedirect.com/science/article/pii/S0167865516300976"},"metadata":{"authorlinks":{"soto, a":"https://asoto.ing.puc.cl/publications/"}},"downloads":0},"search_terms":["proposal","supervised","clustering","dirichlet","process","using","labels","peralta","caro","soto"],"keywords":[],"authorIDs":["32ZR23o2BFySHbtQK","3ear6KFZSRqbj6YeT","4Pq6KLaQ8jKGXHZWH","54578d9a2abc8e9f370004f0","5e126ca5a4cabfdf01000053","5e158f76f1f31adf01000118","5e16174bf67f7dde010003ad","5e1f631ae8f5ddde010000eb","5e1f7182e8f5ddde010001ff","5e26da3642065ede01000066","5e3acefaf2a00cdf010001c8","5e62c3aecb259cde010000f9","5e65830c6e5f4cf3010000e7","5e666dfc46e828de010002c9","6cMBYieMJhf6Nd58M","6w6sGsxYSK2Quk6yZ","7xDcntrrtC62vkWM5","ARw5ReidxxZii9TTZ","BjzM7QpRCG7uCF7Zf","DQ4JRTTWkvKXtCNCp","GbYBJvxugXMriQwbi","HhRoRmBvwWfD4oLyK","JFk6x26H6LZMoht2n","JvArGGu5qM6EvSCvB","LpqQBhFH3PxepH9KY","MT4TkSGzAp69M3dGt","QFECgvB5v2i4j2Qzs","RKv56Kes3h6FwEa55","Rb9TkQ3KkhGAaNyXq","RdND8NxcJDsyZdkcK","SpKJ5YujbHKZnHc4v","TSRdcx4bbYKqcGbDg","W8ogS2GJa6sQKy26c","WTi3X2fT8dzBN5d8b","WfZbctNQYDBaiYW6n","XZny8xuqwfoxzhBCB","Xk2Q5qedS5MFHvjEW","bbARiTJLYS79ZMFbk","cBxsyeZ37EucQeBYK","cFyFQps7W3Sa2Wope","dGRBfr8zhMmbwK6eP","eRLgwkrEk7T7Lmzmf","fMYSCX8RMZap548vv","g6iKCQCFnJgKYYHaP","h2hTcQYuf2PB3oF8t","h83jBvZYJPJGutQrs","jAtuJBcGhng4Lq2Nd","pMoo2gotJcdDPwfrw","q5Zunk5Y2ruhw5vyq","rzNGhqxkbt2MvGY29","uC8ATA8AfngWpYLBq","vMiJzqEKCsBxBEa3v","vQE6iTPpjxpuLip2Z","wQDRsDjhgpMJDGxWX","wbNg79jvDpzX9zHLK","wk86BgRiooBjy323E","zCbPxKnQGgDHiHMWn","zf9HENjsAzdWLMDAu"],"dataSources":["3YPRCmmijLqF4qHXd","sg6yZ29Z2xB5xP79R","sj4fjnZAPkEeYdZqL","m8qFBfFbjk9qWjcmJ","QjT2DEZoWmQYxjHXS"]}