Scalable Bayesian Nonparametric Clustering and Classification

doi:10.1080/10618600.2019.1624366

This content will become publicly available on July 19, 2020

Title: Scalable Bayesian Nonparametric Clustering and Classification

Abstract

We develop a scalable multi-step Monte Carlo algorithm for inference under a large class of nonparametric Bayesian models for clustering and classification. Each step is “embarrassingly parallel” and can be implemented using the same Markov chain Monte Carlo sampler. The simplicity and generality of our approach makes inference for a wide range of Bayesian nonparametric mixture models applicable to large datasets. Specifically, we apply the approach to inference under a product partition model with regression on covariates. We show results for inference with two motivating data sets: a large set of electronic health records (EHR) and a bank telemarketing dataset. We find interesting clusters and competitive classification performance relative to other widely used competing classifiers. Supplementary materials for this article are available online.

Authors:

Ni, Yang ^[1]; Müller, Peter ^[2]; Diesendruck, Maurice ^[3]; Williamson, Sinead ^[4]; Zhu, Yitan ^[5]; Ji, Yuan ^[6]

Texas A & M Univ., College Station, TX (United States). Dept. of Statistics; Univ. of Texas, Austin, TX (United States). Dept. of Statistics and Data Sciences
Univ. of Texas, Austin, TX (United States). Dept. of Mathematics
Univ. of Texas, Austin, TX (United States). Dept. of Statistics and Data Sciences
Univ. of Texas, Austin, TX (United States). Dept. of Information, Risk, and Operations Management
NorthShore Univ. HealthSystem, Evanston, IL (United States). Program for Computational Genomics and Medicine
Univ. of Chicago, IL (United States). Dept. of Public Health Sciences

Publication Date:: 2019-07-19

Research Org.:: Argonne National Lab. (ANL), Argonne, IL (United States)

Sponsoring Org.:: National Institutes of Health (NIH)

OSTI Identifier:: 1566123

Grant/Contract Number:: AC02-06CH11357

Resource Type:: Accepted Manuscript

Journal Name:: Journal of Computational and Graphical Statistics

Additional Journal Information:: Journal Volume: none; Journal Issue: none; Journal ID: ISSN 1061-8600

Publisher:: Taylor & Francis

Country of Publication:: United States

Language:: English

Subject:: 97 MATHEMATICS AND COMPUTING; Electronic health records; non-conjugate models; parallel computing; product partition models

Citation Formats


                    Ni, Yang, Müller, Peter, Diesendruck, Maurice, Williamson, Sinead, Zhu, Yitan, and Ji, Yuan. Scalable Bayesian Nonparametric Clustering and Classification.  United States: N. p., 2019. 
        Web.  doi:10.1080/10618600.2019.1624366.

Copy to clipboard


                    Ni, Yang, Müller, Peter, Diesendruck, Maurice, Williamson, Sinead, Zhu, Yitan, & Ji, Yuan. Scalable Bayesian Nonparametric Clustering and Classification.  United States.  doi:10.1080/10618600.2019.1624366.

Copy to clipboard


                    Ni, Yang, Müller, Peter, Diesendruck, Maurice, Williamson, Sinead, Zhu, Yitan, and Ji, Yuan. Fri .  
        "Scalable Bayesian Nonparametric Clustering and Classification".  United States.  doi:10.1080/10618600.2019.1624366.

Copy to clipboard


                    
@article{osti_1566123,

  title        = {Scalable Bayesian Nonparametric Clustering and Classification},

  author       = {Ni, Yang and Müller, Peter and Diesendruck, Maurice and Williamson, Sinead and Zhu, Yitan and Ji, Yuan},

  abstractNote = {We develop a scalable multi-step Monte Carlo algorithm for inference under a large class of nonparametric Bayesian models for clustering and classification. Each step is “embarrassingly parallel” and can be implemented using the same Markov chain Monte Carlo sampler. The simplicity and generality of our approach makes inference for a wide range of Bayesian nonparametric mixture models applicable to large datasets. Specifically, we apply the approach to inference under a product partition model with regression on covariates. We show results for inference with two motivating data sets: a large set of electronic health records (EHR) and a bank telemarketing dataset. We find interesting clusters and competitive classification performance relative to other widely used competing classifiers. Supplementary materials for this article are available online.},

  doi          = {10.1080/10618600.2019.1624366},

  journal      = {Journal of Computational and Graphical Statistics},

  number       = none,

  volume       = none,

  place        = {United States},

  year         = {2019},

  month        = {7}

}

Copy to clipboard

Journal Article:

Free Publicly Available Full Text

This content will become publicly available on July 19, 2020

Publisher's Version of Record

DOI: 10.1080/10618600.2019.1624366

Other availability

Search WorldCat to find libraries that may hold this journal

Citation Metrics:

Cited by: 1 work

Citation information provided by
Web of Science

Save / Share:

Export Metadata

Save to My Library

Works referenced in this record:

MapReduce: simplified data processing on large clusters
journal, January 2008

Dean, Jeffrey; Ghemawat, Sanjay; Mehta, Brijesh
Communications of the ACM, Vol. 51, Issue 1
DOI: 10.1145/1327452.1327492

Comparing clusterings—an information based distance
journal, May 2007

Meilă, Marina
Journal of Multivariate Analysis, Vol. 98, Issue 5
DOI: 10.1016/j.jmva.2006.11.013

Bayes and big data: the consensus Monte Carlo algorithm
journal, February 2016

Scott, Steven L.; Blocker, Alexander W.; Bonassi, Fernando V.
International Journal of Management Science and Engineering Management, Vol. 11, Issue 2
DOI: 10.1080/17509653.2016.1142191

Two-Stage Metropolis-Hastings for Tall Data
journal, March 2018

Payne, Richard D.; Mallick, Bani K.
Journal of Classification, Vol. 35, Issue 1
DOI: 10.1007/s00357-018-9248-z

Bayesian density estimation and model selection using nonparametric hierarchical mixtures
journal, April 2010

Argiento, Raffaele; Guglielmi, Alessandra; Pievatolo, Antonio
Computational Statistics & Data Analysis, Vol. 54, Issue 4
DOI: 10.1016/j.csda.2009.11.002

Multivariate mixtures of normals with unknown number of components
journal, January 2006

Dellaportas, Petros; Papageorgiou, Ioulia
Statistics and Computing, Vol. 16, Issue 1
DOI: 10.1007/s11222-006-5338-6

Speeding Up MCMC by Efficient Data Subsampling
journal, July 2018

Quiroz, Matias; Kohn, Robert; Villani, Mattias
Journal of the American Statistical Association, Vol. 114, Issue 526
DOI: 10.1080/01621459.2018.1448827

Partition models
journal, January 1990

Hartigan, J. A.
Communications in Statistics - Theory and Methods, Vol. 19, Issue 8
DOI: 10.1080/03610929008830345

Data clustering: 50 years beyond K-means
journal, June 2010

Jain, Anil K.
Pattern Recognition Letters, Vol. 31, Issue 8
DOI: 10.1016/j.patrec.2009.09.011

A data-driven approach to predict the success of bank telemarketing
journal, June 2014

Moro, Sérgio; Cortez, Paulo; Rita, Paulo
Decision Support Systems, Vol. 62
DOI: 10.1016/j.dss.2014.03.001

Markov Chain Sampling Methods for Dirichlet Process Mixture Models
journal, June 2000

Neal, Radford M.
Journal of Computational and Graphical Statistics, Vol. 9, Issue 2
DOI: 10.1080/10618600.2000.10474879

BART: Bayesian additive regression trees
journal, March 2010

Chipman, Hugh A.; George, Edward I.; McCulloch, Robert E.
The Annals of Applied Statistics, Vol. 4, Issue 1
DOI: 10.1214/09-AOAS285

MCMC for Normalized Random Measure Mixture Models
journal, August 2013

Favaro, Stefano; Teh, Yee Whye
Statistical Science, Vol. 28, Issue 3
DOI: 10.1214/13-STS422

Hierarchical Mixture Modeling With Normalized Inverse-Gaussian Priors
journal, December 2005

Lijoi, Antonio; Mena, Ramsés H.; Prünster, Igor
Journal of the American Statistical Association, Vol. 100, Issue 472
DOI: 10.1198/016214505000000132

Bayesian nonparametric classification for spectroscopy data
journal, October 2014

Gutiérrez, Luis; Gutiérrez-Peña, Eduardo; Mena, Ramsés H.
Computational Statistics & Data Analysis, Vol. 78
DOI: 10.1016/j.csda.2014.04.010

Modeling with Normalized Random Measure Mixture Models
journal, August 2013

Barrios, Ernesto; Lijoi, Antonio; Nieto-Barajas, Luis E.
Statistical Science, Vol. 28, Issue 3
DOI: 10.1214/13-STS416

Fitting semiparametric random effects models to large data sets
journal, December 2006

Pennell, M. L.; Dunson, D. B.
Biostatistics, Vol. 8, Issue 4
DOI: 10.1093/biostatistics/kxm008

Controlling the reinforcement in Bayesian non-parametric mixture models
journal, September 2007

Lijoi, Antonio; Mena, Ramsés H.; Prünster, Igor
Journal of the Royal Statistical Society: Series B (Statistical Methodology), Vol. 69, Issue 4
DOI: 10.1111/j.1467-9868.2007.00609.x

Optimal Bayesian estimators for latent variable cluster models
journal, October 2017

Rastelli, Riccardo; Friel, Nial
Statistics and Computing, Vol. 28, Issue 6
DOI: 10.1007/s11222-017-9786-y

Are Gibbs-Type Priors the Most Natural Generalization of the Dirichlet Process?
journal, February 2015

De Blasi, Pierpaolo; Favaro, Stefano; Lijoi, Antonio
IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 37, Issue 2
DOI: 10.1109/TPAMI.2013.217

On a Class of Bayesian Nonparametric Estimates: I. Density Estimates
journal, March 1984

Lo, Albert Y.
The Annals of Statistics, Vol. 12, Issue 1
DOI: 10.1214/aos/1176346412

Heterogeneous reciprocal graphical models: hRGM
journal, October 2017

Ni, Yang; Müller, Peter; Zhu, Yitan
Biometrics, Vol. 74, Issue 2
DOI: 10.1111/biom.12791

Bayesian nonparametric clustering for large data sets
journal, February 2018

Zuanetti, Daiane Aparecida; Müller, Peter; Zhu, Yitan
Statistics and Computing, Vol. 29, Issue 2
DOI: 10.1007/s11222-018-9803-9

Sparse covariance estimation in heterogeneous samples
journal, January 2011

Rodríguez, Abel; Lenkoski, Alex; Dobra, Adrian
Electronic Journal of Statistics, Vol. 5, Issue 0
DOI: 10.1214/11-EJS634

Estimating Normal Means with a Dirichlet Process Prior
journal, March 1994

Escobar, Michael D.
Journal of the American Statistical Association, Vol. 89, Issue 425
DOI: 10.1080/01621459.1994.10476468

The two-parameter Poisson-Dirichlet distribution derived from a stable subordinator
journal, April 1997

Pitman, Jim; Yor, Marc
The Annals of Probability, Vol. 25, Issue 2
DOI: 10.1214/aop/1024404422

Defining Predictive Probability Functions for Species Sampling Models
journal, May 2013

Lee, Jaeyong; Quintana, Fernando A.; Müller, Peter
Statistical Science, Vol. 28, Issue 2
DOI: 10.1214/12-STS407

On Bayesian Analysis of Mixtures with an Unknown Number of Components (with discussion)
journal, November 1997

Richardson, Sylvia.; Green, Peter J.
Journal of the Royal Statistical Society: Series B (Statistical Methodology), Vol. 59, Issue 4
DOI: 10.1111/1467-9868.00095

Fast Bayesian Inference in Dirichlet Process Mixture Models
journal, January 2011

Wang, Lianming; Dunson, David B.
Journal of Computational and Graphical Statistics, Vol. 20, Issue 1
DOI: 10.1198/jcgs.2010.07081

Bayesian Cluster Analysis: Point Estimation and Credible Balls (with Discussion)
journal, June 2018

Wade, Sara; Ghahramani, Zoubin
Bayesian Analysis, Vol. 13, Issue 2
DOI: 10.1214/17-BA1073

Estimating Mixture of Dirichlet Process Models
journal, June 1998

MacEachern, Steven N.; Müller, Peter; Muller, Peter
Journal of Computational and Graphical Statistics, Vol. 7, Issue 2
DOI: 10.2307/1390815

Quantum Support Vector Machine for Big Data Classification
journal, September 2014

Rebentrost, Patrick; Mohseni, Masoud; Lloyd, Seth
Physical Review Letters, Vol. 113, Issue 13
DOI: 10.1103/PhysRevLett.113.130503

Support-vector networks
journal, September 1995

Cortes, Corinna; Vapnik, Vladimir
Machine Learning, Vol. 20, Issue 3
DOI: 10.1007/BF00994018

Bayesian Model-Based Clustering Procedures
journal, September 2007

Lau, John W.; Green, Peter J.
Journal of Computational and Graphical Statistics, Vol. 16, Issue 3
DOI: 10.1198/106186007X238855

Piecewise Approximate Bayesian Computation: fast inference for discretely observed Markov models using a factorised posterior distribution
journal, November 2013

White, S. R.; Kypraios, T.; Preston, S. P.
Statistics and Computing, Vol. 25, Issue 2
DOI: 10.1007/s11222-013-9432-2

Algorithm AS 136: A K-Means Clustering Algorithm
journal, January 1979

Hartigan, J. A.; Wong, M. A.
Applied Statistics, Vol. 28, Issue 1
DOI: 10.2307/2346830

Regularization Paths for Generalized Linear Models via Coordinate Descent
journal, January 2010

Friedman, Jerome; Hastie, Trevor; Tibshirani, Robert
Journal of Statistical Software, Vol. 33, Issue 1
DOI: 10.18637/jss.v033.i01

A Product Partition Model With Regression on Covariates
journal, January 2011

Müller, Peter; Quintana, Fernando; Rosner, Gary L.
Journal of Computational and Graphical Statistics, Vol. 20, Issue 1
DOI: 10.1198/jcgs.2011.09066

Semiparametric Bayesian classification with longitudinal markers
journal, March 2007

Cruz-Mesía, Rolando De la; Quintana, Fernando A.; Müller, Peter
Journal of the Royal Statistical Society: Series C (Applied Statistics), Vol. 56, Issue 2
DOI: 10.1111/j.1467-9876.2007.00569.x

A “Density-Based” Algorithm for Cluster Analysis Using Species Sampling Gaussian Mixture Models
journal, October 2014

Argiento, Raffaele; Cremaschi, Andrea; Guglielmi, Alessandra
Journal of Computational and Graphical Statistics, Vol. 23, Issue 4
DOI: 10.1080/10618600.2013.856796

The Representation of Partition Structures
journal, October 1978

Kingman, J. F. C.
Journal of the London Mathematical Society, Vol. s2-18, Issue 2
DOI: 10.1112/jlms/s2-18.2.374

Sampling the Dirichlet Mixture Model with Slices
journal, January 2007

Walker, Stephen G.
Communications in Statistics - Simulation and Computation, Vol. 36, Issue 1
DOI: 10.1080/03610910601096262

Identifying Mixtures of Mixtures Using Bayesian Estimation
journal, October 2016

Malsiner-Walli, Gertraud; Frühwirth-Schnatter, Sylvia; Grün, Bettina
Journal of Computational and Graphical Statistics, Vol. 26, Issue 2
DOI: 10.1080/10618600.2016.1200472

A scalable bootstrap for massive data
journal, March 2014

Kleiner, Ariel; Talwalkar, Ameet; Sarkar, Purnamrita
Journal of the Royal Statistical Society: Series B (Statistical Methodology), Vol. 76, Issue 4
DOI: 10.1111/rssb.12050

Similar Records in DOE PAGES and OSTI.GOV collections:

A Nonparametric Bayesian Approach For Emission Tomography Reconstruction

Journal Article Barat, Eric ; Dautremer, Thomas - AIP Conference Proceedings

We introduce a PET reconstruction algorithm following a nonparametric Bayesian (NPB) approach. In contrast with Expectation Maximization (EM), the proposed technique does not rely on any space discretization. Namely, the activity distribution--normalized emission intensity of the spatial poisson process--is considered as a spatial probability density and observations are the projections of random emissions whose distribution has to be estimated. This approach is nonparametric in the sense that the quantity of interest belongs to the set of probability measures on R{sup k} (for reconstruction in k-dimensions) and it is Bayesian in the sense that we define a prior directly on thismore »« less
DOI: 10.1063/1.2821285
Unsupervised Group Discovery and LInk Prediction in Relational Datasets: a nonparametric Bayesian approach

Technical Report Koutsourelakis, P

Clustering represents one of the most common statistical procedures and a standard tool for pattern discovery and dimension reduction. Most often the objects to be clustered are described by a set of measurements or observables e.g. the coordinates of the vectors, the attributes of people. In a lot of cases however the available observations appear in the form of links or connections (e.g. communication or transaction networks). This data contains valuable information that can in general be exploited in order to discover groups and better understand the structure of the dataset. Since in most real-world datasets, several of these linksmore »« less
DOI: 10.2172/908093

Full Text Available
Advances in Bayesian Model Based Clustering Using Particle Learning

Technical Report Merl, D M

Recent work by Carvalho, Johannes, Lopes and Polson and Carvalho, Lopes, Polson and Taddy introduced a sequential Monte Carlo (SMC) alternative to traditional iterative Monte Carlo strategies (e.g. MCMC and EM) for Bayesian inference for a large class of dynamic models. The basis of SMC techniques involves representing the underlying inference problem as one of state space estimation, thus giving way to inference via particle filtering. The key insight of Carvalho et al was to construct the sequence of filtering distributions so as to make use of the posterior predictive distribution of the observable, a distribution usually only accessible inmore »« less
DOI: 10.2172/1010386

Full Text Available
Exact and efficient hybrid Monte Carlo algorithm for accelerated Bayesian inference of gene expression models from snapshots of single-cell transcripts

Journal Article Lin, Yen Ting ; Buchler, Nicolas E. - Journal of Chemical Physics

Single cells exhibit a significant amount of variability in transcript levels, which arises from slow, stochastic transitions between gene expression states. Elucidating the nature of these states and understanding how transition rates are affected by different regulatory mechanisms require state-of-the-art methods to infer underlying models of gene expression from single cell data. A Bayesian approach to statistical inference is the most suitable method for model selection and uncertainty quantification of kinetic parameters using small data sets. However, this approach is impractical because current algorithms are too slow to handle typical models of gene expression. To solve this problem, we firstmore »« less
DOI: 10.1063/1.5110503

Full Text Available
astroABC : An Approximate Bayesian Computation Sequential Monte Carlo sampler for cosmological parameter estimation

Journal Article Jennings, E. ; Madigan, M. - Astronomy and Computing

Given the complexity of modern cosmological parameter inference where we arefaced with non-Gaussian data and noise, correlated systematics and multi-probecorrelated data sets, the Approximate Bayesian Computation (ABC) method is apromising alternative to traditional Markov Chain Monte Carlo approaches in thecase where the Likelihood is intractable or unknown. The ABC method is called"Likelihood free" as it avoids explicit evaluation of the Likelihood by using aforward model simulation of the data which can include systematics. Weintroduce astroABC, an open source ABC Sequential Monte Carlo (SMC) sampler forparameter estimation. A key challenge in astrophysics is the efficient use oflarge multi-probe datasets to constrainmore »« less
DOI: 10.1016/j.ascom.2017.01.001

Full Text Available

Other research related to this record: