Nonparametric Mixture of Gaussian Processes with Constraints

James Ross, Jennifer Dy
Proceedings of the 30th International Conference on Machine Learning, PMLR 28(3):1346-1354, 2013.

Abstract

Motivated by the need to identify new and clinically relevant categories of lung disease, we propose a novel clustering with constraints method using a Dirichlet process mixture of Gaussian processes in a variational Bayesian nonparametric framework. We claim that individuals should be grouped according to biological and/or genetic similarity regardless of their level of disease severity; therefore, we introduce a new way of looking at subtyping/clustering by recasting it in terms of discovering associations of individuals to disease trajectories (i.e., grouping individuals based on their similarity in response to environmental and/or disease causing variables). The nonparametric nature of our algorithm allows for learning the unknown number of meaningful trajectories. Additionally, we acknowledge the usefulness of expert guidance by providing for their input using must-link and cannot- link constraints. These constraints are encoded with Markov random fields. We also provide an efficient variational approach for performing inference on our model.

Cite this Paper


BibTeX
@InProceedings{pmlr-v28-ross13a, title = {Nonparametric Mixture of Gaussian Processes with Constraints}, author = {Ross, James and Dy, Jennifer}, booktitle = {Proceedings of the 30th International Conference on Machine Learning}, pages = {1346--1354}, year = {2013}, editor = {Dasgupta, Sanjoy and McAllester, David}, volume = {28}, number = {3}, series = {Proceedings of Machine Learning Research}, address = {Atlanta, Georgia, USA}, month = {17--19 Jun}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v28/ross13a.pdf}, url = {https://proceedings.mlr.press/v28/ross13a.html}, abstract = {Motivated by the need to identify new and clinically relevant categories of lung disease, we propose a novel clustering with constraints method using a Dirichlet process mixture of Gaussian processes in a variational Bayesian nonparametric framework. We claim that individuals should be grouped according to biological and/or genetic similarity regardless of their level of disease severity; therefore, we introduce a new way of looking at subtyping/clustering by recasting it in terms of discovering associations of individuals to disease trajectories (i.e., grouping individuals based on their similarity in response to environmental and/or disease causing variables). The nonparametric nature of our algorithm allows for learning the unknown number of meaningful trajectories. Additionally, we acknowledge the usefulness of expert guidance by providing for their input using must-link and cannot- link constraints. These constraints are encoded with Markov random fields. We also provide an efficient variational approach for performing inference on our model.} }
Endnote
%0 Conference Paper %T Nonparametric Mixture of Gaussian Processes with Constraints %A James Ross %A Jennifer Dy %B Proceedings of the 30th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2013 %E Sanjoy Dasgupta %E David McAllester %F pmlr-v28-ross13a %I PMLR %P 1346--1354 %U https://proceedings.mlr.press/v28/ross13a.html %V 28 %N 3 %X Motivated by the need to identify new and clinically relevant categories of lung disease, we propose a novel clustering with constraints method using a Dirichlet process mixture of Gaussian processes in a variational Bayesian nonparametric framework. We claim that individuals should be grouped according to biological and/or genetic similarity regardless of their level of disease severity; therefore, we introduce a new way of looking at subtyping/clustering by recasting it in terms of discovering associations of individuals to disease trajectories (i.e., grouping individuals based on their similarity in response to environmental and/or disease causing variables). The nonparametric nature of our algorithm allows for learning the unknown number of meaningful trajectories. Additionally, we acknowledge the usefulness of expert guidance by providing for their input using must-link and cannot- link constraints. These constraints are encoded with Markov random fields. We also provide an efficient variational approach for performing inference on our model.
RIS
TY - CPAPER TI - Nonparametric Mixture of Gaussian Processes with Constraints AU - James Ross AU - Jennifer Dy BT - Proceedings of the 30th International Conference on Machine Learning DA - 2013/05/26 ED - Sanjoy Dasgupta ED - David McAllester ID - pmlr-v28-ross13a PB - PMLR DP - Proceedings of Machine Learning Research VL - 28 IS - 3 SP - 1346 EP - 1354 L1 - http://proceedings.mlr.press/v28/ross13a.pdf UR - https://proceedings.mlr.press/v28/ross13a.html AB - Motivated by the need to identify new and clinically relevant categories of lung disease, we propose a novel clustering with constraints method using a Dirichlet process mixture of Gaussian processes in a variational Bayesian nonparametric framework. We claim that individuals should be grouped according to biological and/or genetic similarity regardless of their level of disease severity; therefore, we introduce a new way of looking at subtyping/clustering by recasting it in terms of discovering associations of individuals to disease trajectories (i.e., grouping individuals based on their similarity in response to environmental and/or disease causing variables). The nonparametric nature of our algorithm allows for learning the unknown number of meaningful trajectories. Additionally, we acknowledge the usefulness of expert guidance by providing for their input using must-link and cannot- link constraints. These constraints are encoded with Markov random fields. We also provide an efficient variational approach for performing inference on our model. ER -
APA
Ross, J. & Dy, J.. (2013). Nonparametric Mixture of Gaussian Processes with Constraints. Proceedings of the 30th International Conference on Machine Learning, in Proceedings of Machine Learning Research 28(3):1346-1354 Available from https://proceedings.mlr.press/v28/ross13a.html.

Related Material