International Conference on Artificial Intelligence and Statistics

Page 1

International Conference on

Artificial Intelligence

and Statistics

2013

Program

1

Page 2

Sixteenth International Conference on Artificial Intelligence and Statistics

Monday, April 29

0730 Breakfast – PALM TERRACE 0815 Organizers Welcome – RATTLERS

0830 The Jeopardy! Challenge and Beyond **Invited Talk**

Erik Brown (IBM)

0930 Session I: Learning Theory Chair: Karthik Sridharan (Univ. of Pennsylvania)

Permutation estimation and minimax rates of identifiability ** Notable Paper** Olivier Collier, Arnak Dalalyan (IMAGINE-ENPC / CREST-ENSAE)

Further Optimal Regret Bounds for Thompson Sampling Shipra Agrawal, Navin Goyal, MSR India

1030 Coffee Break – RATTLERS FOYER

1100 Session II: Bayesian Inference I

Chair: Richard Hahn (Univ. of Chicago)

Diagonal Orthant Multinomial Probit Models ** Notable Paper ** James Johndrow (Duke); Kristian Lum (Virgina Tech); David Dunson (Duke)

ODE parameter inference using adaptive gradient matching with Gaussian processes Frank Dondelinger, Biomathematics and Statistics Scotland; Dirk Husmeier, University of Glasgow; Simon Rogers, University of Glasgow; Maurizio Filippone, University of Glasgow

Reconstructing ecological networks with hierarchical Bayesian regression and Mondrian processes Andrej Aderhold, University of St Andrews; Dirk Husmeier, University of Glasgow; V. Anne Smith, University of St Andrews

1230 Afternoon Break – On you own.

1700 Session III: Graphical Models Chair: David Sontag (NYU)

Distributed Learning of Gaussian Graphical Models via Marginal Likelihoods ** Notable Paper ** Zhaoshi Meng Univ. of Michigan; Dennis Wei Univ. of Michigan; Ami Wiesel, Hebrew University; Alfred Hero III, Univ. of Michigan

3

Page 3

Computing the M Most Probable Modes of a Graphical Model Chao Chen, Rutgers University; Vladimir Kolmogorov, IST Austria; Yan Zhu, Rutgers University; Dimitris Metaxas, Rutgers University; Christoph Lampert, IST Austria

Estimating the Partition Function of Graphical Models Using Langevin Importance Sampling Jianzhu Ma, TTIC; Jian Peng,TTIC; Sheng Wang, TTIC; Jinbo Xu, TTIC

1830 Dinner Break – On your own.

2000 Poster Session I – BOUCHON Hors d'oeuvres and cash bar – PALM PATIO

Tuesday, April 30

0730 0830

Breakfast: - PALM TERRACE Oral Session: RATTLERS

0830 Geometric and Topological Inference ** Invited Talk** Larry Wasserman (CMU)

0930 Session IV: Probability Chair: Stephan Clemencon (Telecom ParisTech)

A unifying representation for a class of dependent random measures **Notable Paper** Nicholas Foti (Dartmouth); Sinead Williamson (CMU); Daniel Rockmore (Dartmouth); Joseph Futoma (Dartmouth)

Distribution-Free Distribution Regression Barnabas Poczos, Carnegie Mellon University; Aarti Singh, Carnegie Mellon University; Alessandro Rinaldo, Carnegie Mellon University; Larry Wasserman, IBM

1030 Coffee Break – RATTLERS FOYER 1100 Session V: Sparsity

Chair: Daryl Pregibon (Google)

Sparse Principal Component Analysis for High Dimensional Multivariate Time Series **Notable Paper** Fang Han (JHU); Zhaoran Wang (Princeton); Han Liu (Princeton)

Texture Modeling with Convolutional Spike-and-Slab RBMs and Deep Extensions Heng Luo, Université de Montréal; Pierre Luc Carrier, Université de Montréal; Aaron Courville, Université de Montréal; Yoshua Bengio, Université de Montréal

Detecting Activations over Graphs using Spanning Tree Wavelet Bases James Sharpnack, Carnegie Mellon University; Aarti Singh, Carnegie Mellon University; Akshay Krishnamurthy, CMU

1230 Afternoon Break – On you own.

4

Page 4

1600 Approximative Bayesian Computation (ABC): Computational Advances versus Inferential Uncertainty ** Invited Talk** Christian Robert (Paris)

1700 Session VI: Bayesian Inference II Chair: Emily Fox (Univ. of Washington)

Bayesian learning of joint distributions of objects **Notable Paper** Anjishnu Banerjee (Duke); Jared Murray (Duke); David Dunson (Duke)

Efficient Variational Inference for Gaussian Process Regression Networks Trung Nguyen, ANU and NICTA; Edwin Bonilla*, NICTA and ANU

Structural Expectation Propagation (SEP): Bayesian structure learning for Networks with latent variables Nevena Lazic, Microsoft Research; Christopher Bishop, Microsoft Research ; John Winn, Microsoft Research

1830 Dinner – On your own. 2000 Poster Session II – BOUCHON

Hors d'oeuvres and cash bar – PALM PATIO

Wednesday, May 1

0730 0830

Breakfast – PALM TERRACE Oral Session: RATTLERS

0830 Session VII: Efficient Learning and Inference Chair: Geoff Gordon (CMU)

Faster Training of Structural SVMs with Diverse M-Best Cutting-Planes Abner Guzman-Rivera, University of Illinois; Pushmeet Kohli, Microsoft Research Cambridge; Dhruv Batra, Virginia Tech

Dual Decomposition for Joint Discrete-Continuous Optimization Christopher Zach, Microsoft Research

Nystrom Approximation for Large-Scale Determinantal Processes Raja Hafiz Affandi, University of Pennsylvania; Emily Fox, University of Washington; Ben Taskar, University of Pennsylvania; Alex Kulesza University of Michigan

Supervised Sequential Classification Under Budget Constraints Kirill Trapeznikov, Boston University; Venkatesh Saligrama, Boston University

1030 Coffee Break – RATTLERS FOYER

1100 Session VIII: Learning, Networks and Causality

Chair: Guillaume Bouchard (Xerox)

Statistical Tests for Contagion in Observational Social Network Studies Greg Ver Steeg, Information Sciences Institute; Aram Galstyan, Information Sciences Institute, USC

5

Page 5

Meta-Transportability of Causal Effects: A Formal Approach Elias Bareinboim, UCLA; Judea Pearl, UCLA

Localization and Adaptation in Online Learning Alexander (Sasha) Rakhlin, University of Pennsylvania; Ohad Shamir, Microsoft Research; Karthik Sridharan, University of Pennsylvania

Uncover Topic-Sensitive Information Diffusion Networks Nan Du, Georgia Institute of Technology; Le Song, Georgia Institute of Technology; Hyenkyun Woo, Georgia Institute of Technology; Hongyuan Zha, Georgia Institute of Technology

1300

AISTATS END

6

Page 6

Poster Session I Monday, April 29, 2013 Location: Bouchon

Scoring anomalies: a M-estimation formulation Stéphan Clémençon, Telecom ParisTech; Jérémie Jakubowicz, Telecom Sud Management

Evidence Estimation for Bayesian Partially Observed MRFs Yutian Chen, UC Irvine; Max Welling, "University of California, Irvine

High-dimensional Inference via Lipschitz Sparsity-Yielding Regularizers Zheng Pan, Tsinghua Univ.; Changshui Zhang, Tsinghua Univ.

Learning to Top-K Search using Pairwise Comparisons Brian Eriksson, Technicolor

Stochastic blockmodeling of relational event dynamics Christopher DuBois, UC Irvine; Carter Butts, UC Irvine; Padhraic Smyth, University of California Irvine

Unsupervised Link Selection in Networks Quanquan Gu, CS, UIUC; Charu Aggarwal, IBM Research; Jiawei Han, UIUC

Beyond Sentiment: The Manifold of Human Emotions Seungyeon Kim, Georgia Institute of Technolog; Fuxin Li, Georgia Institute of Technology; Guy Lebanon, Georgia Institute of Technology; Irfan Essa, Georgia Institute of Technology

Greedy Bilateral Sketch, Completion & Smoothing Tianyi Zhou, Universityof Technology Sydney; Dacheng Tao, University of Technology, Sydney

Further Optimal Regret Bounds for Thompson Sampling Shipra Agrawal, MSR India; Navin Goyal, MSR India

Faster Training of Structural SVMs with Diverse M-Best Cutting-Planes Abner Guzman-Rivera, University of Illinois; Pushmeet Kohli, Microsoft Research Cambridge; Dhruv Batra, Virginia Tech

Structural Expectation Propagation (SEP): Bayesian structure learning for networks Nevena Lazic, Christopher Bishop, John Winn, Microsoft Research

DYNA-CARE: Dynamic Cardiac Arrest Risk Estimation Joyce Ho, University of Texas at Austin; Yubin Park, University of Texas at Austin; Carlos Carvalho, University of Texas at Austin; Joydeep Ghosh, University of Texas at Austin

8

Page 7

Competing with an Infinite Set of Models in Reinforcement Learning Phuong Nguyen, Australian National University; Odalric-Ambrym Maillard, Montanuniversität Leoben; Daniil Ryabko, INRIA, Lille; Ronald Ortner, Montanuniversitaet Leoben

Central Limit Theorems for Conditional Markov Chains Mathieu Sinn, IBM Research; Bei Chen, IBM Research – Ireland

Efficient Variational Inference for Gaussian Process Regression Networks Trung Nguyen, ANU and NICTA; Edwin Bonilla, NICTA and ANU

Active Learning for Interactive Visualization Tomoharu Iwata, University of Cambridge; Neil Houlsby, University of Cambridge; Zoubin Ghahramani, University of Cambridge

ODE parameter inference using adaptive gradient matching with Gaussian processes Frank Dondelinger, Biomathematics and Statistics Scotland; Dirk Husmeier, University of Glasgow; Simon Rogers, University of Glasgow; Maurizio Filippone, University of Glasgow

Exact Learning of Bounded Tree-width Bayesian Networks Janne Korhonen, University of Helsinki; Pekka Parviainen

Completeness Results for Lifted Variable Elimination Nima Taghipour, KU Leuven; Daan Fierens, KU LEUVEN; Guy Van den Broeck, UCLA; Jesse Davis, KU LEUVEN; Hendrik Blockeel, KU LEUVEN

Fast Near-GRID Gaussian Process Regression Yuancheng Luo, University of Maryland; Ramani Duraiswami, University of Maryland

Convex Collective Matrix Factorization Guillaume Bouchard, "Xerox Research Centre, Europe"; Dawei Yin, Lehigh University; Shengbo Guo, Samsung Research America

Meta-Transportability of Causal Effects: A Formal Approach Elias Bareinboim, UCLA; Judea Pearl, UCLA

Why Steiner-tree type algorithms work for community detection Mung Chiang, Princeton University; Henry Lam, Boston University; Zhenming Liu, Princeton University; Harold Poor, Princeton University

Clustering Oligarchies Margareta Ackerman, Caltech; Shai Ben David, ; David Loker, University of Waterloo; Sivan Sabato, Microsoft Research

Structure Learning of Mixed Graphical Models Jason Lee, Computational Math & Engineeri; Trevor Hastie, Stanford University

A Simple Criterion for Controlling Selection Bias Eunice Yuh-Jie Chen, UCLA; Judea Pearl, UCLA

9

Page 8

Clustered Support Vector Machine Quanquan Gu, CS, UIUC; Jiawei Han, UIUC

A Competitive Test for Uniformity of Monotone Distributions Jayadev Acharya, University of California, San Diego; Ashkan Jafarpour, Univ. of California, San Diego; Alon Orlitsky, University of California, San Diego; Ananda Suresh, University of California, San Diego

Deep Gaussian Processes Andreas Damianou, University of Sheffield; Neil Lawrence, University of Sheffield

Permutation estimation and minimax rates of identifiability Olivier Collier, IMAGINE-ENPC / CREST-ENSAE; Arnak Dalalyan, Ecole des Ponts ParisTech

Bayesian Structure Learning for Functional Neuroimaging Oluwasanmi Koyejo, UT Austin; Mijung Park, UT Austin; Russell Poldrack, UT Austin; Joydeep Ghosh, UT Austin; Jonathan Pillow, The University of Texas at Austin

Dual Decomposition for Joint Discrete-Continuous Optimization Christopher Zach, Microsoft Research

Distribution-Free Distribution Regression Barnabas Poczos, Carnegie Mellon University; Aarti Singh, Carnegie Mellon University; Alessandro Rinaldo, Carnegie Mellon University; Larry Wasserman, Carnegie Mellon University

A Last-Step Regression Algorithm for Non-Stationary Online Learning Edward Moroshko, Technion; Koby Crammer, Technion University

Efficiently Sampling Probabilistic Programs via Program Analysis Arun Chaganty, Aditya Nori, Sriram Rajamani, Microsoft Research India

10

Page 9

Poster Session II Tuesday, April 30, 2013 Location: Bouchon

On the Asymptotic Optimality of Maximum Margin Bayesian Networks Sebastian Tschiatschek, TU Graz; Franz Pernkopf, TU Graz

Ultrahigh Dimensional Feature Screening via RKHS Embeddings Krishnakumar Balasubramanian, Gatech; Bharath Sriperumbudur, Cambridge University ; Guy Lebanon, Georgia Institute of Technology

Data-driven covariate selection for nonparametric estimation of causal effects Doris Entner, Patrik Hoyer, University of Helsinki; Peter Spirtes, CMU

Mixed LICORS: A Nonparametric Algorithm for Predictive State Reconstruction Georg Goerg, Carnegie Mellon University; Cosma Shalizi, Carnegie Mellon University

Thompson Sampling in Switching Environments with Bayesian Online Change Detection Joseph Mellor, University of Manchester; Jonathan Shapiro, University of Manchester

Collapsed Variational Bayesian Inference for Hidden Markov Models Pengyu Wang, University of Oxford; Phil Blunsom, University of Oxford

Supervised Sequential Classification Under Budget Constraints Kirill Trapeznikov, Boston University; Venkatesh Saligrama, Boston University

Computing the M Most Probable Modes of a Graphical Model Chao Chen, Rutgers University; Vladimir Kolmogorov, IST Austria; Yan Zhu, Rutgers University; Dimitris Metaxas, Rutgers University; Christoph Lampert, IST Austria

Estimating the Partition Function of Graphical Models Using Langevin Importance Sampling Jianzhu Ma, TTIC; Jian Peng, TTIC; Sheng Wang, TTIC; Jinbo Xu, TTIC

Random Projections for Support Vector Machines Saurabh Paul, Rensselaer Polytechnic Inst; Christos Boutsidis, IBM; Malik Magdon-Ismail, ; Petros Drineas, RPI

A unifying representation for a class of dependent random measures Nicholas Foti, Dartmouth College; Sinead Williamson, Carnegie Mellon University; Daniel Rockmore, Dartmouth College; Joseph Futoma, Dartmouth College

Dynamic Copula Networks for Modeling Real-valued Time Series Elad Eban, Hebrew University; Gideon Rothschild, Hebrew University; Adi Mizrahi, Hebrew University; Israel Nelken, Hebrew University; Gal Elidan, Hebrew University

11

Page 10

A Parallel, Block Greedy Method for Sparse Inverse Covariance Estimation for Ultra-high Dimensions Prabhanjan Kambadur, IBM TJ Watson Research Center; Aurelie Lozano,

A recursive estimate for the predictive likelihood in a topic model James Scott, Jason Baldridge, University of Texas at Austin

Nystrom Approximation for Large-Scale Determinantal Processes Raja Hafiz Affandi, University of Pennsylvania; Emily Fox, Ben Taskar, Alex Kulesza, University of Pennsylvania

A simple sketching algorithm for entropy estimation over streaming data Ioana Cosma, University of Ottawa; Peter Clifford, University of Oxford

Detecting Activations over Graphs using Spanning Tree Wavelet Bases James Sharpnack, Carnegie Mellon University; Aarti Singh, Carnegie Mellon University; Akshay Krishnamurthy, CMU

Learning Social Infectivity in Sparse Low-rank Networks Using Multi-dimensional Hawkes Processes Ke Zhou, Georgia Institute of Technolog; Le Song, Georgia Institute of Technology; Hongyuan Zha, Georgia Institute of Technology

Changepoint Detection over Graphs with the Spectral Scan Statistic James Sharpnack, Carnegie Mellon University; Aarti Singh, Carnegie Mellon University; Alessandro Rinaldo, Carnegie Mellon University

Statistical Tests for Contagion in Observational Social Network Studies Greg Ver Steeg, Information Sciences Institute; Aram Galstyan, Information Sciences Institute, USC

Diagnonal Orthant Multinomial Probit Models James Johndrow, Duke University; Kristian Lum, Virginia Tech; David Dunson, Duke University

Reconstructing ecological networks with hierarchical Bayesian regression and Mondrian processes Andrej Aderhold, University of St Andrews; Dirk Husmeier, University of Glasgow; V. Anne Smith, University of St Andrews

Bayesian learning of joint distributions of objects Anjishnu Banerjee, Duke University; Jared Murray, Duke University; David Dunson, Duke University

Consensus Ranking with Signed Permutations Raman Arora, TTIC; Marina Meila, University of Washington

Sparse Principal Component Analysis for High Dimensional Multivariate Time Series Zhaoran Wang, Princeton University; Fang Han, Johns Hopkins University; Han Liu, Princeton Univ.

Texture Modeling with Convolutional Spike-and-Slab RBMs and Deep Extensions Heng Luo, Université de Montréal; Pierre Luc Carrier, Université de Montréal; Aaron Courville, Université de Montréal; Yoshua Bengio, Université de Montréal

12

Page 11

Block Regularized Lasso for Multivariate Multi-Response Linear Regression Weiguang Wang, Syracuse University; Yingbin Liang, Syracuse University; Eric Xing, Carnegie Mellon University

Predictive Correlation Screening: Application to Two-stage Predictor Design in High Dimension Hamed Firouzi, University of Michigan; Alfred Hero III, University of Michigan

Distributed Learning of Gaussian Graphical Models via Marginal Likelihoods Zhaoshi Meng, University of Michigan; Dennis Wei, University of Michigan; Ami Wiesel, The Hebrew University of Jerusalem; Alfred Hero III, University of Michigan

Dynamic Scaled Sampling for Deterministic Constraints Lei Li, UC Berkeley; Bharath Ramsundar, Stanford; Stuart Russell, UC Berkeley

Localization and Adaptation in Online Learning Alexander (Sasha) Rakhlin, University of Pennsylvania; Ohad Shamir, Karthik Sridharan, University of Pennsylvania

Recursive Karcher Expectation Estimators And Geometric Law of Large Numbers Hesamoddin Salehian, University of Florida; Guang Cheng, ; Jeffrey Ho, UFL; Baba Vemuri, University of Florida

Distributed and Adaptive Darting Monte Carlo through Regenerations Sungjin Ahn, UCI; Yutian Chen, UC Irvine; Max Welling, University of Amsterdam

Uncover Topic-Sensitive Information Diffusion Networks Nan Du, GATECH; Le Song, Hyenkyun Woo, Hongyuan Zha, Georgia Institute of Technology

Learning Markov Networks With Arithmetic Circuits Daniel Lowd, University of Oregon; Amirmohammad Rooshenas, University of Oregon

Bethe Bounds and Approximating the Global Optimum Adrian Weller, Columbia University; Tony Jebara, Columbia University

13

Page 12

OFF

ICES

OFF

ICES

91

48

91

47

91

46

91

45

91

44

91

43

91

42

91

41

91

35

91

33

91

31

91

29

91

27

91

25

91

23

91

21

91

36

91

34

91

32

91

30

91

28

91

26

91

24

91

22

91

15

91

13

91

11

91

09

91

07

91

05

SAG

UA

RO

CH

OLL

A

91

01

91

16

91

14

91

12

91

10

91

06

91

04

91

02

51

06

51

07

51

08

51

09

51

10

51

11

51

12

51

13

41

13

41

14

51

14

41

15

41

12

41

11

41

10

41

08

41

09

41

07

41

06

81

19

81

18

81

07 8

10

8 81

03 81

04

61

10

61

09

61

07 6

10

8 61

03 61

04

71

11

71

10

71

06

71

05

71

02

71

01

11

33

11

32

11

37

11

36

811781168115811481138112

8111

8110

812481258126812781287120

710771096106

71126113

71147115

6115

7113

61126114

6111

81308129

81328133

8123

8131

81218122

8120

21062109

21022101

3104

7104

6101

2105

3108

7108

61053103

7103

6102

211231133112

2115

3115

2113

3111

2114

3114

113011291120

1135

1123

1131

1121

1134

1122

3107

7119

711871177116

611661176118611961205101

5102

510351045105

311631173118311931204101

4102

410341044105

21

11

21

10

21

07

21

08

21

03

21

04

11

19

11

18

11

061

10

511

021

10

1

31

10

31

09

31

063

10

531

023

10

1

112411251126

11271128212021192118

21172116

8105

8106 1107

8102

8101

11081103

1104

8109

111711161115111411131112

1111

1110

1109

INTE

RIO

R W

ALK

WAY

SPU

BLI

C F

AC

ILIT

IES

EXTE

RIO

R W

ALK

WAY

SPO

OLS

& F

OU

NTA

INS

VEN

DIN

G M

AC

HIN

ESIC

E M

AC

HIN

ESEL

EVAT

ORS

STA

IRS

PUB

LIC

RES

TRO

OM

SI

VE

PARK

ING

PARK

ING

PARK

ING

PARKING PARKING

PARKINGPARKING

PARK

ING

PARK

ING

PAR

AD

ISE

PARK

TEN

NIS

CO

URT

S

JACK RABBIT ROAD

SCO

TTSD

ALE

RO

AD

TO O

LD T

OW

N S

CO

TTSD

ALE

NO

RTH

SWIM

MIN

G P

OO

L

WHIR

LPOOLSO

UTH

POO

L B

AR

FIRE

PLA

CE

WHIR

LPOOL

FOY

ERLO

BB

Y

REG

ISTR

ATIO

N

ENTR

Y

GIF

TSH

OPLOGGIA

PALM

PAT

IO

CH

AM

BER

SR

ATTL

ERS

PAYSON

BO

UC

HO

N

EN F

UEG

ORE

STA

UR

AN

T

BUSI

NES

SC

ENTE

R

EAST

PARK

WES

TPA

RK

RIN

GFO

UN

TAIN

CA

CTU

SG

ARD

EN

SOU

THSW

IMM

ING

POO

L

CH

ILD

REN

’SPL

AYG

RO

UN

D

PUTT

ING

GRE

EN

THE

FORU

M

HEA

LTH

CLU

B

SAN

CARL

OS

GRA

ND

BALL

ROO

M

CHA

PARR

AL

MEE

TIN

G/

CON

VEN

TIO

N/

CATE

RIN

GO

FFIC

E

PRES

COTT

SON

ORA

RIO

VERD

E

DIR

ECTO

R’S

BOA

RDRO

OM

EXEC

UTI

VE B

OA

RDRO

OM

SED

ON

A

PALO

MA

S

CORO

NA

DO

SIER

RAAV

IS

FIR

ST F

LOO

R

CA

BA

NA

BA

R

V E EI

II

I

II

I

E

I