{"type":"ScholarlyArticle","title":"Entangled Kernels","authors":[{"name":"Riikka Huusari","slug":"riikka-huusari","given":"Riikka","family":"Huusari","affiliation":"Aix-Marseille University , CNRS , LIS , Marseille , France"},{"name":"Hachem Kadri","slug":"hachem-kadri","given":"Hachem","family":"Kadri","affiliation":"Aix-Marseille University , CNRS , LIS , Marseille , France"}],"abstract":"We consider the problem of operator-valued kernel learning and \r\ninvestigate the possibility of going beyond the well-known separable \r\nkernels. Borrowing tools and concepts from the field of quantum computing, such \r\nas partial trace and entanglement, we propose a new view on \r\noperator-valued kernels and define a general family of kernels that \r\nencompasses previously known operator-valued kernels, including \r\nseparable and transformable kernels. Within this framework, we introduce another novel class of \r\noperator-valued kernels called entangled kernels that are not \r\nseparable. We propose an efficient two-step algorithm for this framework, where the\r\n entangled kernel is learned based on a novel extension of kernel \r\nalignment to operator-valued kernels. The utility of the algorithm is illustrated on both artificial and real data.","doi":"10.24963/ijcai.2019/358","url":"https://ijcai.org/proceedings/2019/358","pdf":"https://ijcai.org/proceedings/2019/0358.pdf","markdown":"https://ijcai.org/proceedings/2019/358.md","bibtex":"https://ijcai.org/proceedings/2019/bibtex/358","volume":{"slug":"2019","title":"Proceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence","short_title":"IJCAI 2019","url":"https://ijcai.org/proceedings/2019/","year":2019,"location":"Macao","dates":"10-16 August 2019","editors":"Sarit Kraus","isbn":"978-0-9992411-4-1","publisher":"International Joint Conferences on Artificial Intelligence Organization","doi_prefix":"10.24963/ijcai.2019"},"section":"Machine Learning A-L","track":"Main track","number":358,"pages":{"first":2578,"last":2584},"keywords":["Machine Learning: Kernel Methods","Machine Learning: Multi-instance;Multi-label;Multi-view learning"],"videos":[],"references":[{"position":1,"raw":"[Alvarez et al., 2012] Mauricio A. Alvarez, Lorenzo Rosasco, Neil D. Lawrence, et al. Kernels for vector- valued functions: A review. Foundations and Trends R in Machine Learning, 4(3):195-266, 2012.","title":"Kernels for vectorvalued functions: A review","authors":"Mauricio A Alvarez, Lorenzo Rosasco, Neil D Lawrence","year":"2012","venue":"Foundations and Trends R in Machine Learning","doi":null,"url":null,"internal_url":null},{"position":2,"raw":"[Attal, 2015] Stephane Attal. Lectures in quantum noise the- ory. Lecture 6: Quantum channels, 2015.","title":"Lectures in quantum noise theory. Lecture 6: Quantum channels","authors":"Stephane Attal","year":"2015","venue":null,"doi":null,"url":null,"internal_url":null},{"position":3,"raw":"[Bengtsson and Życzkowski, 2017] Ingemar Bengtsson and Karol Życzkowski. Geometry of quantum states: an intro- duction to quantum entanglement. Cambridge university press, 2017.","title":"Geometry of quantum states: an introduction to quantum entanglement","authors":"Życzkowski Bengtsson","year":"2017","venue":null,"doi":null,"url":null,"internal_url":null},{"position":4,"raw":"[Brault et al., 2016] Romain Brault, Markus Heinonen, and Florence Buc. Random Fourier features for operator- valued kernels. In ACML, pages 110-125, 2016.","title":"Random Fourier features for operatorvalued kernels","authors":"Romain Brault, Markus Heinonen, Florence Buc","year":"2016","venue":"ACML","doi":null,"url":null,"internal_url":null},{"position":5,"raw":"[Brouard et al., 2016] Céline Brouard, Marie Szafranski, and Florence d'Alché Buc. Input output kernel regression: Supervised and semi-supervised structured output predic- tion with operator-valued kernels. JMLR, 17(176):1-48, 2016.","title":"Input output kernel regression: Supervised and semi-supervised structured output prediction with operator-valued kernels","authors":"Brouard","year":"2016","venue":"JMLR","doi":null,"url":null,"internal_url":null},{"position":6,"raw":"[Caponnetto et al., 2008] Andrea Caponnetto, Charles A. Micchelli, Massimiliano Pontil, and Yiming Ying. Uni- versal multi-task kernels. JMLR, 9(Jul):1615-1646, 2008.","title":"Universal multi-task kernels","authors":"Andrea Caponnetto, Charles A Micchelli, Massimiliano Pontil, Yiming Ying","year":"2008","venue":"JMLR","doi":null,"url":null,"internal_url":null},{"position":7,"raw":"[Carmeli et al., 2010] Claudio Carmeli, Ernesto De Vito, Alessandro Toigo, and Veronica Umanita. Vector valued reproducing kernel hilbert spaces and universality. Analy- sis and Applications, 08(01):19-61, 2010.","title":"Vector valued reproducing kernel hilbert spaces and universality","authors":"Claudio Carmeli, Ernesto De Vito, Alessandro Toigo, Veronica Umanita","year":"2010","venue":"Analysis and Applications","doi":null,"url":null,"internal_url":null},{"position":8,"raw":"[Choi, 1975] Man-Duen Choi. Completely positive linear maps on complex matrices. Linear algebra and its ap- plications, 10(3):285-290, 1975.","title":"Completely positive linear maps on complex matrices","authors":"Man-Duen Choi, Choi","year":"1975","venue":"Linear algebra and its applications","doi":null,"url":null,"internal_url":null},{"position":9,"raw":"[Ciliberto et al., 2015] Carlo Ciliberto, Youssef Mroueh, Tomaso Poggio, and Lorenzo Rosasco. Convex learning of multiple tasks and their structure. In ICML, 2015.","title":"Convex learning of multiple tasks and their structure","authors":"Carlo Ciliberto, Youssef Mroueh, Tomaso Poggio, Lorenzo Rosasco","year":"2015","venue":"ICML","doi":null,"url":null,"internal_url":null},{"position":10,"raw":"[Cortes et al., 2010] Corinna Cortes, Mehryar Mohri, and Afshin Rostamizadeh. Two-stage learning kernel algo- rithms. In ICML, pages 239-246, 2010.","title":"Two-stage learning kernel algorithms","authors":"Corinna Cortes, Mehryar Mohri, Afshin Rostamizadeh","year":"2010","venue":"ICML","doi":null,"url":null,"internal_url":null},{"position":11,"raw":"[Cristianini et al., 2002] Nello Cristianini, John Shawe- Taylor, Andre Elisseeff, and Jaz S Kandola. On kernel- target alignment. In NIPS, pages 367-373, 2002.","title":"On kerneltarget alignment","authors":"Nello Cristianini, John Shawe-Taylor, Andre Elisseeff, Jaz S Kandola","year":"2002","venue":"NIPS","doi":null,"url":null,"internal_url":null},{"position":12,"raw":"[Dinuzzo and Fukumizu, 2011] Francesco Dinuzzo and Kenji Fukumizu. Learning low-rank output kernels. In ACML, 2011.","title":"Learning low-rank output kernels","authors":"Francesco Dinuzzo, Kenji Fukumizu","year":"2011","venue":"ACML","doi":null,"url":null,"internal_url":null},{"position":13,"raw":"[Dinuzzo et al., 2011] Francesco Dinuzzo, Cheng S. Ong, Gianluigi Pillonetto, and Peter V Gehler. Learning output kernels with block coordinate descent. In ICML, 2011.","title":"Learning output kernels with block coordinate descent","authors":"Francesco Dinuzzo, S Cheng, Gianluigi Ong, Peter V Pillonetto, Gehler","year":"2011","venue":"ICML","doi":null,"url":null,"internal_url":null},{"position":14,"raw":"[Evgeniou et al., 2005] Theodoros Evgeniou, Charles A. Micchelli, and Massimiliano Pontil. Learning multiple tasks with kernel methods. JMLR, 6:615-637, 2005.","title":"Learning multiple tasks with kernel methods","authors":"Theodoros Evgeniou, Charles A Micchelli, Massimiliano Pontil","year":"2005","venue":"JMLR","doi":null,"url":null,"internal_url":null},{"position":15,"raw":"[Filipiak et al., 2018] Katarzyna Filipiak, Daniel Klein, and Erika Vojtková. The properties of partial trace and block trace operators of partitioned matrices. 33, 2018.","title":"The properties of partial trace and block trace operators of partitioned matrices","authors":"Katarzyna Filipiak, Daniel Klein, Erika Vojtková","year":"2018","venue":null,"doi":null,"url":null,"internal_url":null},{"position":16,"raw":"[Gregorová et al., 2017] Magda Gregorová, Alexandros Kalousis, and Stéphane Marchand-Maillet. Forecast- ing and granger modelling with non-linear dynamical dependencies. In ECML-PKDD, 2017. [","title":"Forecasting and granger modelling with non-linear dynamical dependencies","authors":"Magda Gregorová, Alexandros Kalousis, Stéphane Marchand-Maillet","year":"2017","venue":"ECML-PKDD","doi":null,"url":null,"internal_url":null},{"position":17,"raw":"[Horodecki et al., 2009] Ryszard Horodecki, Paweł Horodecki, Michał Horodecki, and Karol Horodecki. Quantum entanglement. Reviews of modern physics, 81(2):865, 2009.","title":"Quantum entanglement","authors":"Ryszard Horodecki, Paweł Horodecki, Michał Horodecki, Karol Horodecki","year":"2009","venue":"Reviews of modern physics","doi":null,"url":null,"internal_url":null},{"position":18,"raw":"[Jawanpuria et al., 2015] Pratik Jawanpuria, Maksim Lapin, Matthias Hein, and Bernt Schiele. Efficient output kernel learning for multiple tasks. In NIPS, 2015.","title":"Efficient output kernel learning for multiple tasks","authors":"Pratik Jawanpuria, Maksim Lapin, Matthias Hein, Bernt Schiele","year":"2015","venue":"NIPS","doi":null,"url":null,"internal_url":null},{"position":19,"raw":"[Kadri et al., 2012] Hachem Kadri, Alain Rakotomamonjy, Philippe Preux, and Francis R Bach. Multiple operator- valued kernel learning. In NIPS, 2012.","title":"Multiple operatorvalued kernel learning","authors":"Hachem Kadri, Alain Rakotomamonjy, Philippe Preux, Francis R Bach","year":"2012","venue":"NIPS","doi":null,"url":null,"internal_url":null},{"position":20,"raw":"[Kadri et al., 2016] Hachem Kadri, Emmanuel Duflos, Philippe Preux, Stéphane Canu, Alain Rakotomamonjy, and Julien Audiffren. Operator-valued kernels for learning from functional response data. JMLR, 16:1-54, 2016.","title":"Operator-valued kernels for learning from functional response data","authors":"Hachem Kadri, Emmanuel Duflos, Philippe Preux, Stéphane Canu, Alain Rakotomamonjy, Julien Audiffren","year":"2016","venue":"JMLR","doi":null,"url":null,"internal_url":null},{"position":21,"raw":"[Kraus, 1983] Karl Kraus. States, effects and operations: fundamental notions of quantum theory. Springer, 1983.","title":"States, effects and operations: fundamental notions of quantum theory","authors":"Karl Kraus","year":"1983","venue":null,"doi":null,"url":null,"internal_url":null},{"position":22,"raw":"[Lim et al., 2015] Néhémy Lim, Florence d'Alché Buc, Cédric Auliac, and George Michailidis. Operator-valued kernel-based vector autoregressive models for network in- ference. Machine learning, 99(3):489-513, 2015.","title":"Operator-valued kernel-based vector autoregressive models for network inference","authors":"Lim","year":"2015","venue":"Machine learning","doi":null,"url":null,"internal_url":null},{"position":23,"raw":"[Micchelli and Pontil, 2005] Charles A. Micchelli and Mas- similiano Pontil. On learning vector-valued functions. Neural Computation, 17:177-204, 2005.","title":"Micchelli and Massimiliano Pontil. On learning vector-valued functions","authors":"A Charles","year":"2005","venue":"Neural Computation","doi":null,"url":null,"internal_url":null},{"position":24,"raw":"[Minh et al., 2016] Ha Quang Minh, Loris Bazzani, and Vit- torio Murino. A unifying framework in vector-valued re- producing kernel hilbert spaces for manifold regularization and co-regularized multi-view learning. JMLR, 17(25):1- 72, 2016.","title":"Ha Quang Minh, Loris Bazzani, and Vittorio Murino. A unifying framework in vector-valued reproducing kernel hilbert spaces for manifold regularization and co-regularized multi-view learning","authors":"Minh","year":"2016","venue":"JMLR","doi":null,"url":null,"internal_url":null},{"position":25,"raw":"[Minh, 2016] Ha Quang Minh. Operator-valued Bochner theorem, Fourier feature maps for operator-valued ker- nels, and vector-valued learning. arXiv preprint arXiv:1608.05639, 2016.","title":"Operator-valued Bochner theorem, Fourier feature maps for operator-valued kernels, and vector-valued learning","authors":"Ha Quang, Minh","year":"2016","venue":null,"doi":null,"url":null,"internal_url":null},{"position":26,"raw":"[Mintert et al., 2009] F Mintert, C Viviescas, and A Buch- leitner. Basic concepts of entangled states. In Entangle- ment and Decoherence, pages 61-86. Springer, 2009.","title":"Basic concepts of entangled states","authors":"Mintert, Viviescas, Buchleitner","year":"2009","venue":"Entanglement and Decoherence","doi":null,"url":null,"internal_url":null},{"position":27,"raw":"[Petersen and Pedersen, 2008] Kaare Brandt Petersen and Michael Syskind Pedersen. The matrix cookbook. Techni- cal University of Denmark, 7(15):510, 2008.","title":"The matrix cookbook","authors":"Pedersen Petersen, Kaare Brandt Petersen, Michael Syskind Pedersen","year":"2008","venue":"Technical University of Denmark","doi":null,"url":null,"internal_url":null},{"position":28,"raw":"[Rahimi and Recht, 2008] Ali Rahimi and Benjamin Recht. Random features for large-scale kernel machines. In NIPS, pages 1177-1184, 2008.","title":"Random features for large-scale kernel machines","authors":"Recht Rahimi, Ali Rahimi, Benjamin Recht","year":"2008","venue":"NIPS","doi":null,"url":null,"internal_url":null},{"position":29,"raw":"[Rieffel and Polak, 2011] Eleanor G. Rieffel and Wolf- gang H. Polak. Quantum computing: A gentle introduc- tion. MIT Press, 2011.","title":"Quantum computing: A gentle introduction","authors":"Eleanor G Rieffel, Wolfgang H Polak","year":"2011","venue":null,"doi":null,"url":null,"internal_url":null},{"position":30,"raw":"[Sangnier et al., 2016] Maxime Sangnier, Olivier Fercoq, and Florence d'Alché Buc. Joint quantile regression in vector-valued rkhss. In NIPS, pages 3693-3701, 2016.","title":"Joint quantile regression in vector-valued rkhss","authors":"Sangnier","year":"2016","venue":"NIPS","doi":null,"url":null,"internal_url":null},{"position":31,"raw":"[Sindhwani et al., 2013] Vikas Sindhwani, Minh Ha Quang, and Aurélie C. Lozano. Scalable matrix-valued kernel learning for high-dimensional nonlinear multivariate re- gression and granger causality. In UAI, 2013.","title":"Scalable matrix-valued kernel learning for high-dimensional nonlinear multivariate regression and granger causality","authors":"Vikas Sindhwani, Minh Ha Quang, Aurélie C Lozano","year":"2013","venue":"UAI","doi":null,"url":null,"internal_url":null},{"position":32,"raw":"[Williams and Seeger, 2001] Christopher KI Williams and Matthias Seeger. Using the nyström method to speed up kernel machines. In NIPS, pages 682-688, 2001.","title":"Using the nyström method to speed up kernel machines","authors":"K I Christopher, Matthias Williams, Seeger","year":"2001","venue":"NIPS","doi":null,"url":null,"internal_url":null}],"license_note":"Copyright International Joint Conferences on Artificial Intelligence."}