Annual Report 2018-2019 · Bayesian nonparametric framework for high dimensional problems. Visiting Researchers The Department of Statistics and Data Sciences hosted five visiting

http://stat.utexas.edu

Annual Report

2018-2019

http://stat.utexas.edu

The Department of Statistics and Data Sciences is an academic unit

housed in the College of Natural Sciences that supports the statistical and

data science needs of The University of Texas at Austin campus.

Our Mission

Our mission is to be a world-class center for statistical science that

advances scholarship across The University of Texas at Austin and

prepares graduates to flourish in an information-rich world.

Our Vision The Department of Statistics and Data Sciences will become a vibrant

community of scholars that has a strong intellectual identity and is widely

recognized for excellence in interdisciplinary research and high impact

educational programs.

Our Core Values

• Commitment to education

• Excellence in research

• Broad engagement with science and society

• Collegiality and integrity

Department of Statistics and Data Sciences i

From the Interim Chair It is a bittersweet moment as I pen this letter, for it marks the end

of my term as the interim chair of the department. For as much as

I enjoyed leading the continued growth of the Department of

Statistics and Data Sciences (SDS), I am happy to welcome Dr.

Kate Calder as the new department chair, starting in September

2019. Kate is an outstanding scholar and leader who will be joining

us from The Ohio State University. I expect that our reputation as

a leader in statistics and data sciences will be internationally

recognized under her guidance with her bold vision for growth.

As excited as I am to look forward, the past academic year 2018–

2019 has its share of significant events and noteworthy achievements. First, in addition to

recruiting Kate as department chair, we successfully recruited Dr. Antonio Linero, a stellar

researcher with a research focus in Bayesian nonparametrics for causal inference. We recruited

Tony from Florida State University. He will be joining us as an Assistant Professor in August

2019. The combined recruitment of Kate and Tony has effectively doubled our number of 100%

SDS faculty. We congratulate Dr. James Scott for promotion to Professor, making his formal

rank match his recognition and standing in the scientific community.

Notable achievements for our instructional faculty include Dr. Kristin Harvey being named the

inaugural Director of Undergraduate Studies and the very deserved promotion of three Assistant

Professors of Instruction to Associate Professor of Instruction: Dr. Kristin Harvey, Sally

Ragsdale, and Dr. Lindsey Smith. The instructional team also welcomed three new faculty, Dr.

Choudur Lakshminarayan, Dr. Nathaniel Woodward, and Jason Rarick. Combined, their

contribution resulted in 21 courses having first-rate instruction from committed educators.

As further evidence that SDS faculty are making waves not only at the university but also on the

international platform, four faculty were recognized with prestigious awards. Dr. Kristin Harvey

received the President’s Associates Teaching Excellence Award, Sally Ragsdale received a

CNS Teaching Excellence Award, Dr. Cory Zigler received the 2019 Rothman Epidemiology

Prize, and I was awarded the Zellner Medal.

Achievements were not limited to faculty, many students completed SDS programs: 9 MS in

Statistics graduates, 17 Portfolio in Applied Statistical Modeling graduates, 2 Portfolio in

Scientific Computation graduates, 21 Certificate in Applied Statistical Modeling graduates, and 8

Certificate in Scientific Computation and Data Sciences graduates. Three PhD in Statistics

students were awarded University Fellowships to support their research as they near

graduation: Su Chen, Maurice Diesendruck, and Carlos Pagani Zanini.

Additionally, the consulting team and the annual Summer Statistics Institute continued to

support training in statistics and data science for the university community and beyond.

One of the greatest challenges facing the university this year was the transition of the HR

system to Workday in November 2018. The roll-out has been less than smooth but has

presented the staff the opportunity to meet this challenge head-on and renew efforts to work

across colleges and units to ensure that everyone gets paid. I would like to especially recognize

our new Senior Administrative Associate Victoria Obioma for her continued efforts in mastering

Workday. As the initiator of processes that were once straightforward but are now complicated

and time-intensive, Victoria, and the rest of the administrative staff, has my sincere gratitude.

Department of Statistics and Data Sciences ii

I would like to thank the dedicated faculty and staff of the department for all the hard work

during this critical year of transition. I am also grateful for the support of Deans Paul Goldbart

and David Vanden Bout and their team in the CNS dean’s office, who continue their unwavering

support of our department.

I’m looking forward to the next and upcoming years as we expect incredible growth with

significant impact under the new leadership of Dr. Kate Calder.

Peter Müller

Department of Statistics and Data Sciences iii

Table of Contents

I. SDS Highlights in 2018–2019 ...................................................................................... 1

II. Unit Reports

A. Instruction .............................................................................................................. 5

B. Professional Education ........................................................................................... 8

C. Consulting ............................................................................................................. 9

D. Graduate Degree Programs ................................................................................ 11

E. Portfolio and Certificate Programs ....................................................................... 13

F. SDS Seminar Series ............................................................................................ 17

G. Summer Statistics Institute .................................................................................. 18

H. Corporate Sponsorships ....................................................................................... 19

I. Grants .................................................................................................................... 20

J. Development ......................................................................................................... 23

III. Appendices

A. Organizational Chart ............................................................................................ 24

B. SDS Core Faculty ................................................................................................. 25

C. SDS Seminar Series Speakers ........................................................................... 32

Department of Statistics and Data Sciences 1

I. HIGHLIGHTS IN 2018–2019

New Chair Announcement In January 2019, Dean Paul Goldbart announced that the Department of Statistics and Data

Sciences will welcome Dr. Catherine Calder as Professor and Department Chair for Fall 2019.

Dr. Calder is currently a professor of statistics and co-director of the Mathematical Biosciences

Institute at The Ohio State University.

New Personnel The Department of Statistics and Data Sciences welcomed six new personnel in 2018–2019:

• Dr. Corwin Zigler joined SDS in September 2018 as an Associate Professor. Most

recently from Harvard School of Public Health, Cory received his PhD from the

Department of Biostatistics at the UCLA Fielding School of Public Health. Cory’s

research focuses on statistical methodology to confront the challenges of complex

observational studies. Areas of focus include Bayesian methods, causal inference,

comparative effectiveness research, spatial statistics, and environmental health data

science.

• Dr. Choudur Lakshminarayan joined SDS faculty in Fall 2018 as an Adjunct Assistant

Professor teaching SDS 387 Linear Models. Since 2013, Lakshminarayan has been an

integral part of the annual Summer Statistics Institute roster, teaching big data topics;

most recently offering “Big Data Analytics: Theory and Methods” in May 2019.

• Jason Rarick joined SDS as Lecturer in Fall 2018 teaching SDS 302 Data Analysis for the Health Sciences. Prior to joining SDS, Jason received an M.A. in Educational

Psychology at The University of Texas at Austin before going on to New York University

to pursue a PhD in Applied Psychology.

• Dr. Nathaniel Woodward joined SDS as a Lecturer in Fall 2018 teaching SDS 328M Biostatistics. Prior to joining SDS, Nathaniel earned his PhD in Educational Psychology

in 2018 and M.S. in Statistics in 2016 from The University of Texas at Austin. Nathaniel

taught his first SDS course in 2017 as an Assistant Instructor.

• Dr. Robert Lunde joined SDS in Fall 2018 as a Post-Doctoral Fellow supervised by Dr.

Purnamrita Sarkar. Prior to joining SDS, Robert obtained his PhD in Statistics at

Carnegie Mellon University. Robert’s research interests include time series, networks,

and the bootstrap.

• Victoria Obioma joined SDS as the Senior Administrative Associate in September 2018.

Prior to joining SDS, Victoria worked at the Office of Sponsored Projects and earned a

B.A. in Radio, Television and Film from The University of Texas at Austin. She

participated in multiple internships in Los Angeles and Austin including an internship at

Black Entertainment Television (BET).

New Director of Undergraduate Studies Dr. Kristin Harvey was appointed as the Director of Undergraduate Studies for the department in

the fall of 2018. Dr. Harvey had served the department in many ways over the past six years,

including acting as course coordinator for two of undergraduate foundation statistics courses.

She has been responsible for evaluating course equivalency forms from other departments,


coordinating departmental extended time testing, helping with course planning and TA

allotment. Additionally, she serves on the CNS Course and Curriculum committee and on the

Faculty Advisory committees for the Quantitative Reasoning Flag and the Texas Success

Initiative. She has also served on the CNS 21st Century Task force for the Century Curriculum

Planning and Implementation Task force, the CNS Service Course Committee, the STEM

Council under Dr. David Laude, and the CNS Non-Tenure Track Committee. She was recently

elected to be the Vice Chair of Educational Policy Committee of the Faculty Council and is the

co-chair of the Task Force on Student Feedback (revising course instructor surveys)

commissioned by the Faculty Council and the Provost.

In addition to her service, Dr. Harvey has won several teaching awards, including the

President’s Associates Teaching Excellence Award. Dr. Harvey serves as a valuable resource

to the department for innovative teaching approaches as well as serving as a liaison between

SDS and the other departments and colleges we serve through our undergraduate foundation

courses.

Faculty Awards Dr. Kristin Harvey was recognized with the President’s Associates Teaching Excellence Award.

The award recognizes the university's educational innovators who demonstrate exceptional

undergraduate teaching in the core curriculum, including signature courses, and who engage

with curriculum reform and educational innovation.

Dr. Peter Mueller was awarded the Zellner Medal from the International Society for Bayesian

Analysis. The purpose of the Zellner Medal is to recognize ISBA members who have rendered

exceptional and distinguished service to ISBA over an extended period of time, and whose

contributions have had an impact on the society beyond the time of their incumbency.

Sally Ragsdale was honored with a CNS Teaching Excellence Award. This award recognizes

and honors the committed and innovative educators who help students learn a demanding

curriculum and excel in the classroom.

Dr. Cory Zigler received the 2019 Rothman Epidemiology Prize. This prize is given annually for

the best paper published in the Epidemiology Journal in the preceding year. The selection

criteria are importance, originality, clarity of thought, and excellence in writing. Dr. Zigler’s

winning paper, titled “Impact of National Ambient Air Quality Standards Nonattainment

Designations on Particulate Pollution and Health,” appeared in the March 2018 issue.

Faculty Promotions The President announced the promotion of four Department of Statistics and Data Sciences

faculty members in February 2019. The promotions are expected to be approved by the UT

System Board of Regents in August and be effective September 1, 2019.

• Dr. James Scott: promotion to full Professor

• Dr. Kristin Harvey: promotion to Associate Professor of Instruction

• Sally Ragsdale: promotion to Associate Professor of Instruction

• Dr. Lindsey Smith: promotion to Associate Professor of Instruction


Faculty Search The Department of Statistics and Data Sciences conducted a year-long search for a new junior

faculty position that sits 100% in SDS. After reviewing 151 applications, six applicants were

invited for on-campus interviews. A faculty position was offered to one candidate who will be

joining SDS in Fall 2019:

• Antonio Linero comes to UT from the Department of Statistics at Florida State University

where he is currently an Assistant Professor. Dr. Linero earned his PhD in Statistics from

the University of Florida. His research is broadly focused on developing flexible

Bayesian methods. His work has focused on developing appropriate Bayesian methods

for complex longitudinal data, as well as developing model selection tools within the

Bayesian nonparametric framework for high dimensional problems.

Visiting Researchers The Department of Statistics and Data Sciences hosted five visiting researchers this year:

• Nicolás Kuschinski comes from Centro de Investigación en Matemáticas, A.C., where he

is pursuing a Doctorate in Probability and Statistics. His research interests are in

Bayesian Statistics. During his residency at UT, he worked with Peter Mueller.

• Jack Jewson comes from the University of Warwick where he is a fourth year student on

the OxWaSP CDT. His research looks at Bayesian methods for making more robust

decisions under misspecified models and is working under the supervision of Jim Smith

(Warwick). and Chris Holmes (Oxford). During his residency at UT, he worked with Dr.

Stephen Walker.

• Cristiano Villa comes from the University of Kent, where he is a Senior Lecturer in

Statistics. His research is Objective Bayesian analysis, Bayesian model selection and

change point analysis, and Bayesian nonparametric modelling for “big data” problems

and cyber-security. During his residency at UT, he worked with Dr. Stephen Walker.

• He Zhou comes from Monash University, where he is concluding his PhD in Faculty

Information Technology. His research is in Statistical machine learning, especially on

Bayesian/deep/neural generative models for discrete data, e.g. texts and relational

networks, with applications in natural language processing, social network analysis, and

collaborative filtering. During his residency at UT, he worked with Dr. Mingyuan Zhou.

• Ying (Rachel) Wang comes from the University of Sydney, Australia, where she is a

Lecturer. She earned her PhD in Statistics from the University of California–Berkeley.

Her research interests are in theoretical and applied statistical problems arising from

genomic studies, in particular, the reconstruction and modeling of gene networks. During

her residency at UT she worked with Dr. Purnamrita Sarkar.

Community Outreach: Thinkery’s “Get Your Game On”

SDS participated in the Thinkery’s May adult-only event, “Thinkery21: Get Your Game On.” The

Thinkery in Austin is a private non-profit organization that offers children hands-on, interactive

STEAM (science, technology, engineering, arts, and math) exhibits and programs. The general

target age of the Thinkery entrant is newborn through eleven years old. However, adults want to

have hands-on fun, too; thus, the monthly “Thinkery21” event that provides adults the chance to

explore and learn.


May’s Thinkery21 theme was “Get Your Game On,” which celebrated the love of games. We

decided that the classic Monty Hall problem would be simple to play but challenging enough to

keep people interested. Dr. Nathaniel Woodward developed an interactive computer app that

allowed participants to play the game in real time and make the same decisions participants in

the show would make. The classic setup involves three doors, two choices, one car, and some

goats. The goal of the game is to win the car. Participants make their first door selection using

the app and then also execute their “keep” or “switch” choice after seeing what’s behind the first

door. Dr. Woodward and Steven Hernandez hosted the game and provided explanations about

the two strategies and why one (switch) is statistically better than the other. As an incentive for

playing the game, winning participants were given a baked good provided by Baked from

Scratch ATX. Organizers of the event said the booth was one of the most popular of the night!

SDS Community Drives: The Department of Statistics and Data Sciences is proud to contribute to the UT community in

ways that reach beyond our front door. This year the department held donation drives for

Orange Santa and the Food Pantry and Career Closet.

SDS participated in Orange Santa for the 4th consecutive year in December 2018. The initiative

to participate was originally proposed by Dr. Kristin Harvey in 2014, and has become an end-of-

the year tradition. Orange Santa began in 1994 as a program designed to foster a more caring

university environment by providing toys for UT families in need during the holiday season. The

department gathered toys through the month of November and participated in the Holiday sing-

a-long in December hosted by Orange Santa.

SDS held a food pantry and career closet drive during April 2019 to benefit UT Outpost. UT

Outpost is hosted by UT’s Student Emergency Services in the Office of the Dean of Students

and addresses food insecurity and professional clothing needs of UT students.

2018–2019 Committee Service Staff Member Committee Position

Kristin Harvey

General Faculty Standing Committee:

Educational Policy Committee Member

Faculty Advisory Committee for Texas Success

Initiative Member

CNS Service Courses Committee Member

Faculty Advisory Committee for Quantitative

Reasoning Flag Member

Course and Curriculum Committee, College of

Natural Sciences Member

Stephanie Tomlinson UT Association of Research Administrators Member

T32 Collaborative Learning Group Leader

Vicki Keller

General Faculty Standing Committee:

Responsibilities, Rights, & Welfare of Graduate

Student Academic Employees Committee

Member


II. UNIT REPORTS

A. INSTRUCTION

The demand for statistics and scientific computation courses continues to be strong, particularly

at the foundation course-level. SDS has responded by adding sections of in-demand courses to

meet the needs of the diverse UT student population. SDS now offers more than 100

undergraduate and graduate courses each year, ranging from introductory data analysis and

undergraduate honors statistics, to advanced graduate topics in Bayesian modeling and

computational statistics.

Enrollment Trends

SDS faculty taught 71 undergraduate courses and 41 graduate courses to a total 5719 students

in 2018–2019.


Campus Usage SDS courses serve a broad cross-section of students across campus.

Enrollment in undergraduate SDS courses during 2018–2019 is shown by college:

Enrollment in graduate SDS courses during 2018–2019 is shown by college:


Statistics Online Support (SOS) Dr. Kristin Harvey and Ms. Sally Ragsdale were awarded a Curriculum Innovation Grant from

the Center for Teaching and Learning in Spring 2015 that funded the creation of the Statistics

Online Support (SOS) resource platform (freely available at https://sites.utexas.edu/sos).

Designed for student researchers who have collected data but need guidance selecting and

performing statistical analyses, SOS provides on-demand content and tutorials that guide users

to independently select the data analysis method appropriate for answering their research

questions. Specifically, Dr. Harvey and Ms. Ragsdale created a website that provides users a

structured decision tree based platform where users follow embedded links that guide them to

the types of statistical methods appropriate for their situation and tutorial videos on how to carry

out those analyses.

There continues to be a high level of usage for the site. In May 2019 alone, there were more

than 4,282 new active users accessing the site and 389 returning users from dozens of

countries spread across six continents. In the past year, there have been 2,921 returning

visitors to the site and 39,342 new visitors. Of those returning, 1,800 people return to the site

three or more times with 800 people returning to the site a week or more after their first visit.

This indicates that SOS is a reliable resource for data analysis that many people return to use

multiple times. The most popular tutorials are on Pearson Correlation and Linear Regression

with 1,497 pageviews in May 2019. The next most popular tutorial was Frequency Tables, Pie

Charts, and Bar Charts with 1,440 pageviews.

MOOC: For the 2018–2019 academic year, Dr. Michael Mahometa continued to offer the department’s

Massively Open Online Course (MOOC), “Foundations of Data Analysis” (FoDA), through the

edX platform. This current version was an asynchronous, archived release, allowing students to

move through the course at their own speed. The course was offered in two parts: one for

Descriptive Statistics and Visualization, and another for Inferential Statistics—each containing

six weeks of instruction. FoDA includes 12 total weeks of instruction, with over 40 instructional

content videos, over 20 R tutorial videos, over 120 direct feedback videos for incorrect answers,

and a scaffolded learning structure that reliably guides students through the analysis process,

while increasing independence. Dr. Mahometa plans on offering a FoDA with Instructor

Feedback in the Spring of 2020.

OnRamps Beginning Fall 2013, Dr. Michael Mahometa, and Ms. Sally Ragsdale worked alongside Dr.

Cathy Stacy, and with the Center for Teaching and Learning, to transform SDS 302 Data

Analysis for the Health Sciences into a course for The University of Texas OnRamps initiative.

This initiative continues to serve a wide variety of Texas High School students, offering dual

enrollment for college credit in SDS 302. In recent years, SDS has seen an explosion of

enrollment in the OnRamps Statistics dual enrollment participation. This past academic year

(2018-2019), under the leadership of Dr. Mahometa and Ms. Rebecca Lyle, OnRamps Statistics

was offered in 59 schools (up from 30 last year), with 1983 total enrollments (up from 792 last

year). Of the students enrolled this year, 48% were eligible for course credit, with over half

(54%) accepting their UT SDS 302 college credit.

Noteworthy:

• OnRamps Statistics will be undergoing a redesign in syllabus and teaching strategies for

a Fall 2020 rollout. Planning includes incorporating a greater emphasis on simulation

through acclaimed R packages (via project MOSAIC), and the inclusion of a final student

project.


B. PROFESSIONAL EDUCATION

As the Department of Statistics and Data Sciences grows, so does the need to provide a visible

presence to a larger community. Currently, outreach is through our annual Summer Statistics

Institute.

As the Director of Professional Education and Consulting, Dr. Michael J. Mahometa is tasked

with expanding the department’s professional education offerings. Over the next few years, he

plans to roll out a variety of new educational offerings. These offerings include, but are not

limited, to:

• Statistical Consulting Blog. The Statistical Consulting blog, written by the

department’s Statistical Consultants, will focus on common pain points the Statistical

Consulting Group sees their clients experiencing. This blog will complement the

Statistical Consulting FAQs, and offer a more hands-on asynchronous educational

offering for those both in and out of The University of Texas at Austin community.

• Online Software Short Course offerings. With the continued success of Statistical

Consulting’s Software Short Courses, SDS will create a series of Online Short Courses.

The first series will include Dr. Mahometa’s R and RStudio courses focusing on the

widely used dplyr and ggplot2 packages. These courses are planned to be hosted

through edX, the Department’s partner for the Foundations of Data Analysis MOOC.

• Option III Master’s Program. The Department is also investigating the delivery of a 30-

hour Master’s Program in Data Science and Statistics. This program will serve a broad

population of data consumers, focusing on core concepts in statistics and applications to

a wide variety of fields. We are happy to have the support of the College of Natural

Sciences’ Office of Graduate Education in this exciting endeavor.

In addition to the above listed ventures, Professional Education at the department has been

involved in several new bootcamps, including Biomedical Engineering’s REU (Research

Experiences for Undergraduates) 2019 program and the College of Natural Sciences’ Graduate

Student Data Science training. The department’s role in these bootcamps included:

• Biomedical Engineering’s REU: The creation of 6 hours of material to focus on

Statistical Thinking and the Scientific Method. This included discussions of crafting a

research question, refining in to a hypothesis statement, and applying an appropriate

statistical test. Students examined simulated null distributions to better understand p-

values, and collaboratively worked through a test and graphical summary of a common

engineering failure.

• CNS’ Graduate Student Data Science Training: The application of 12 hours of

material twice this past academic year, covering the common uses of R to import,

transform data, to create new variables from data; graphical investigations of data using

ggplot2 and summary statistics via dplyr; analysis of data including common t-tests,

ANOVA, regression, logistic regression, and multilevel models. This Bootcamp was

offered as part of the CNS Graduate Education Professional Development in both the

Fall of 2018 and Spring of 2019, with continued plans to be offered and expanded upon.

The Department’s Professional Education arm of the Statistical Consulting Group will

continue its involvement with the offering of more advanced courses in future offerings.


C. CONSULTING

SDS provides free statistical consulting services to students, faculty, and staff. Clients receive

assistance in planning and interpreting analyses, working with statistical software, developing

research study designs, and learning how to better organize and manipulate their data. Faculty

may also request contract consulting services for more in-depth analyses. In addition, SDS

offers a collection of highly-subscribed instructional short courses in various statistical methods

and software packages for a nominal fee.

Free Consulting Free consulting services continued to be in high demand in 2018-2019. SDS consultants

provided 649.25 hours of free consulting to UT faculty, staff, and students, for an average of

13.53 hours per academic week. Student clients used 81.94% of these hours, while faculty and

staff used the remaining 18.06%. Of clients responding to the post-consultation survey, 99.32%

reported that their most recent consult was either “very good” or “excellent.” Of the responding

clients, 98.58% reported that their research results will be of higher quality, and 99.3% felt more

confident about doing research because of the assistance provided by SDS Consulting

Services.

Contract Consulting SDS provided 83 hours of contract consulting to UT faculty and organizational units and Dell

Medical School clinical faculty in 2018–2019. Contract consulting clients generally have more

complex data analysis needs and pay a fee for these services. SDS consultants typically

provide a detailed reporting of results suitable for publication. Contract consulting clients during

the 2018–2019 academic year included but was not limited to: faculty and staff from the College

of Liberal Arts, Department of Linguistics (40%);, the College of Liberal Arts, Department of

Government (28%); the School of Nursing (12%); and Dell Medical School, Department of

Dermatology (19%).


Short Courses In 2018–2019, SDS taught 25 short courses to a total of 522 students, faculty, and staff. Most of

these three-hour courses cover specific software, such as R, Python, SQL, or Stata. In addition,

SDS also continued offering the Topics in Statistics series taught by the statistical consulting

staff.

Overall, SDS’s software and topic short courses saw high enrollment and positive participant

feedback in 2018–2019. Of the participants completing the post-course survey, 94.6% said they

would recommend the course they took to others.

Custom Short Courses In addition to the yearly offerings of Software Short Courses, SDS Consulting Services offered

four Custom Short Courses. Two courses were created as part of CNS’s Data Science training

for graduate students. The third, was an introduction to statistical thinking and the scientific

method for the Biomedical Engineering REU program. This course was offered to 12 REU

undergraduates during their first week of a summer study program. The fourth Custom Short

Course was presented to the Study Abroad International Office staff, covering introductory and

advanced topics in Excel.

Noteworthy:

• SDS Consulting services continued to be in high demand across the University, with

over 99% of all 2017–2018 available appointments filled and an average wait time

exceeding 8 days. �

• SDS short courses continued to have high enrollment and positive participant feedback

in 2018–2019. �

• Evan Stein, doctoral student in Psychology, served as the Graduate Student consultant

during spring 2019. Post-graduation Evan will be hired as a part-time consultant.


D. SDS GRADUATE DEGREE PROGRAMS

PhD in Statistics The PhD in Statistics program welcomed its sixth cohort of seven students in August 2018. The

program received 153 completed applications for Fall 2019 admission. Ten offers of admission

were made, resulting in an expected yield of five students.

Admissions Data: PhD Statistics Fall

2013

Fall

2014

Fall

2015

Fall

2016

Fall

2017

Fall

2018

Fall

2019

Applied 79 104 130 138 125 129 153

Admitted 13 13 15 15 20 19 10

Selectivity (%) 16.46 12.5 11.5 10.9 16 14.7 6.5

Enrolled 5 6 8 5 7 7 5*

Yield (%) 38.46 46.15 53.33 33.33 35.00 36.84 50.00

* Expected

PhD Student Progress

Academic Year

Enrollment

Count

Candidacy

Count Graduates

2018–2019* 30 15 -

2017–2018 25 14 2

2016–2017 20 8 2

2015–2016 16 3 -

2014–2015 11 - -

2013–2014 5 - -

TOTAL 4 *Enrollment and graduation numbers current through Spring 2019 Master’s of Science in Statistics The M.S. in Statistics program only accepted applications from current UT graduate students

who are applying to earn the degree concurrently with a doctoral degree in another discipline.

Enrollment Trends Enrollment Count

Academic Year M.S. Only M.S./PhD Graduates

2018–2019* 4 32 9

2017–2018 9 22 9

2016–2017 7 31 13

2015–2016 10 37 19

2014–2015 28 25 18

2013–2014 22 18 12

2012–2013 30 17 19

2011–2012 28 10 11

2010–2011 20 5 7

TOTAL 117 *Enrollment and graduation numbers current through Spring 2019


Noteworthy:

• Three PhD in Statistics students entered candidacy during 2018–2019: Ciara Nugent,

Giorgio Paulon, and Spencer Woody.

• Three PhD in Statistics students have been awarded Fellowships from the UT Graduate

School:

o Su Chen, University Graduate Continuing Fellowship for 2019-2020

o Maurice Diesendruck, University of Texas at Austin Graduate School Summer

2019 Fellowship

o Carlos Pagani Zanini, University of Texas at Austin Graduate School Summer

2019 Fellowship

• Seven PhD students have made a total of 12 presentations of their research at national

or international conferences during 2018–2019.

• Seven PhD students have a total of nine publications either accepted or published

during 2018–2019.


E. PORTFOLIO AND CERTIFICATE PROGRAMS

SDS provides four unique opportunities for students seeking to develop competencies in

statistical modeling or scientific computation. The Portfolio in Applied Statistical Modeling and

the Portfolio in Scientific Computation are 12-credit programs available to graduate students

interested in strengthening and applying these skills to their research area. The Certificate in

Applied Statistical Modeling and the Certificate in Scientific Computation & Data Sciences are

similar 18-hour programs available to undergraduate students.

Enrollment Trends

Portfolio in Applied Statistical Modeling Since its inception in Fall 2009, 287 students have been admitted into the program and 134

students have completed the portfolio requirements.

Academic Year Enrollment Count Completion Count

2018–2019* 40 17

2017–2018 54 18

2016–2017 40 13

2015–2016 40 17

2014–2015 39 15

2013–2014 47 16

2012–2013 54 15

2011–2012 43 11

2010–2011 24 11

2009–2010 9 1


Participating students come from 11 colleges and schools across campus.

Student Participation Snapshot College/School # students currently enrolled # students completed

Cockrell School of Engineering 13 31

College of Communication — 8

College of Education 8 21

College of Liberal Arts 6 34

College of Natural Sciences 4 15

Jackson School of Geosciences 2 3

LBJ School of Public Affairs 1 16

McCombs School of Business — 2

School of Nursing — 1

School of Pharmacy 2 1

School of Social Work 4 2

TOTAL 40 134


Portfolio in Scientific Computation Since its inception in Fall 2010, 38 students have been admitted to the program and 12 students

have completed the portfolio requirements.


2018–2019* 8 2

2017–2018 11 2

2016–2017 10 1

2015–2016 5 1

2014–2015 15 3

2013–2014 12 1

2012–2013 9 2

2011–2012 7 —

2010–2011 7 —


Participating students come from three colleges and schools across campus.




School of Pharmacy — 1

TOTAL 8 12

Certificate in Applied Statistical Modeling Since its inception in Fall 2013, 476 students have enrolled and 85 students have completed the

program.


2018–2019* 255 21

2017–2018 139 29

2016–2017 97 25

2015–2016 56 8

2014–2015 27 1

2013–2014 9 1



Participating students come from six colleges and schools across campus.



College of Communication 1 —

College of Education 2 —



McCombs School of Business 74 14

Undeclared 1 —

TOTAL 255 85

Certificate in Scientific Computation & Data Sciences Since its inception in Fall 2009, 404 students have been admitted into the program, and 58

students have completed the program.


2018–2019* 120 8

2017–2018 99 9

2016–2017 84 8

2015–2016 69 7

2014–2015 69 4

2013–2014 79 5

2012–2013 71 6

2011–2012 51 6

2010–2011 27 4

2009–2010 19 1


Participating students come from seven colleges and schools across campus.



College of Communication 1 --

College of Fine Arts 1 --



Jackson School of Geosciences 2 1

McCombs School of Business 11 3

Undeclared 1 --

TOTAL 120 58


Noteworthy: • To meet the increasing demand of both certificate programs, automatic enrollment

began in Spring 2019. With this change, both programs have increased student

participation since AY 2017–2018.

• Current enrollment in the Certificate in Applied Statistical Modeling is 83% higher and

current enrollment in the Certificate in Scientific Computation & Data Sciences is 21%

higher, compared to last year.


F. SDS SEMINAR SERIES

SDS Seminar Series Since Fall 2011, the SDS Seminar Series has made a vital contribution to the intellectual,

cultural, and scholarly environment at The University of Texas at Austin for students, faculty,

and the wider community. The lecture series provides participants with the opportunity to hear

from leading scholars and experts who work in different applied areas, including business,

biology, medicine, computer modeling, and economics. Each talk is free of charge and open to

the public.

Dr. Abhra Sarkar chaired the SDS Seminar Series committee that made the final selection of

the invited speakers. PhD student Ciara Nugent acted as the student host and liaison for the

invited speakers. The Fall 2018 SDS Seminar Series featured fourteen speakers with an

average attendance of 25 people. The Spring 2019 SDS Seminar Series featured 10 speakers

with an average attendance of 20. (Please see Appendix C for program details.)

Noteworthy: • PhD in Statistics students in candidacy present a seminar on their research during their

fourth year in the program. Six students presented during the 2018–2019 season: o Su Chen o Evan Ott o Carlos Pagani Zanini o Mengie Wang

o Matteo Vestrucci

o Mingzhang Yin


G. SUMMER STATISTICS INSTITUTE

The 12th annual UT Summer Statistics Institute (SSI) was held May 28–31, 2019, in

collaboration with Academic Technology Support in the Patton Hall (RLP) building. SSI provides

a unique hands-on opportunity for participants to acquire valuable skills directly from experts in

the field. Participants joined the Institute from across the country, coming from as far away as

Hawaii and New York, illustrating the continued popularity SSI nationwide. SSI offered 24

twelve-hour courses designed to appeal to a broad range of students, faculty, staff, and the

public. New courses this year included “Intermediate Excel for Personal and Professional Use”

with Maliki Ghossainy, “Scalable Machine Learning: Methods and Tools” with Weijia Xu, and

“Statistical Methods for Categorical Data—Logistical Regression and Beyond” with Dr. Abhra

Sarkar.

The 2019 SSI had enrollment of 394 participants filling 513 seats with 24 participating

instructors from 11 departments. This year’s attendance breakdown was: 25% UT students,

15% UT faculty and staff, 12% non-UT students, and 48% non-UT other (e.g., private industry,

state departments, and non-profits).

The 2019 SSI brought in a gross income of $194,430 with a projected total net profit of $86,369,

a 17% increase over 2018.

Noteworthy: • SDS implemented Cvent as the new SSI registration system, bringing multiple

innovations and ease of use for SSI participants and SDS staff. • 60% of SSI enrollees came from outside The University of Texas at Austin. A first in

program history. • 100% of SSI survey respondents said that the course will benefit their professional

opportunities, while 96% said that they would recommend SSI to others.


H. CORPORATE PARTNERSHIPS Dr. Bindu Viswanathan met with representatives from the following companies to discuss

potential collaborative efforts:

• Social Solutions

• Tyson Foods

• USAA

• Merck & Co.

• Tata Consultancy

During spring 2019, students in the consulting seminar class worked on projects from Tyson

Foods, from Dr. Valerie Danesh from the School of Nursing, and Dr. Farya Phillips from the

School of Social Work.

.


I. GRANTS

During 2018–2019, SDS assisted faculty with the submission of more than five individual

faculty-driven research proposals and three continuations to the following organizations:

• National Institutes of Health (NIH)

• National Science Foundation (NSF)

and in collaboration with:

• NorthShore University Health System

• Harvard University

SDS had 5 active grants and contracts this fiscal year:

Faculty Active Grants & Contracts

Lauren Meyers,

Professor

• Program Director: Predoctoral Training in Biomedical Big Data Science, NIH, ($1,018,560)

Peter Müller,

Professor

• Sub-award PI: Bayesian Inference for Tumor Heterogeneity with Next-Generation Sequencing Data, NIH R01 ($45,433)

Purnamrita Sarkar,

Assistant Professor

• PI: Collaborative Research: Inference for Network Models with Covariates: Leveraging Local Information for Statistically and Computationally Efficient Estimation of Global Parameters,

NSF, ($160,000)

Stephen Walker,

Professor

• PI: Bayesian Nonparametric Time Dependent Models, ConTex,

($33,500)

Corwin Zigler,

Associate Professor

• PI: ACE Center: Regional Air Pollution Mixtures: The Past and Future Impacts of Emission Controls and Climate Change on Air Quality and Health, EPA, ($54,208)

BIG DATA TO KNOWLEDGE (BD2K): T32 Predoctoral Training in Biomedical Big Data 2018–2019 was the 3

rd year of the five-year T32 (NIH) predoctoral training grant in biomedical

big data for a total project amount of $1,018,560 with $197,756 total costs awarded in this 3rd

year. The BD2K project was developed to deepen predoctoral students’ knowledge in three key

areas: biology, statistics, and computer science in order to provide the holistic training

necessary to meaningfully mine biomedical big data.


Starting in Fall 2018 the Executive Committee selected three new trainees to join the two

continuing trainees from the previous year:

• Mackenzie Johnson (Ecology, Evolution, and Behavior)

• Karl Muller (Neuroscience)

• Jennifer Starling (Statistics and Data Sciences)

Continuing trainees:

• Anthony Dutcher (Neuroscience)

• Ciara Nugent (Statistics and Data Sciences)

New trainees began their training by taking the three core courses and the seminar/workshop

course. Core courses included “CSE 380: Tools and Techniques for Computer Science,” “BIO

382K: Introduction to Biology for Data Science,” and “SDS 385: Statistical Models for Big Data.”

Continuing trainees were on rotation, mentored by participating faculty members in areas

outside the trainees’ graduate programs.

Two new trainees have been selected to start the program in Fall 2019.

Dr. Lauren Meyers acted as Program Director for 2018–2019.

Noteworthy:

• SDS graduate students were encouraged to apply for the NIH Ruth L. Kirschstein

National Research Service Award Individual Predoctoral Fellowship (F31). Students also

attended the F31 application workshop provided by the university. One predoctoral

student applied for external funding from this opportunity, results pending.

• Two trainees presented papers and posters in the past year:

o Ciara Nugent (2nd

year trainee)

§ “Inferring treatment effects for unplanned subgroups using multiple

studies,” Bayesian Causal Inference Workshop, June 2019

o Jennifer E. Starling (1st year trainee)

§ “Bayesian Causal Forests with Targeted Smoothing for Identification of

Heterogeneous Treatment Effects: An Application to Family Planning,”

2019 Atlantic Causal Inference Conference, May 2019

§ “BART with Targeted Smoothing: An Application to Patient-Specific Still

Birth Risk,” 2019 International Chinese Statistical Association

Symposium, June 2019

§ “Targeted Smooth Bayesian Causal Forests for Time-Varying

Heterogeneous Treatment Effects,” 2019 International Chinese Statistical

Association Symposium, June 2019

• Two trainees published papers in this past year:

o Karl S. Muller (1st year trainee)

§ Zhang, R., Liu, Z., Zhang, L., Whritner, J. A., Muller, K. S., Hayhoe, M.

M., & Ballard, D. H. (2018). Agile: Learning attention from human for

visuomotor tasks. In Proceedings of the European Conference on Computer Vision (ECCV) (pp. 663-679).

o Jennifer Starling (1st year trainee)

§ Starling, J. E., Murray, J. S., Lohr, P. A., Aiken, A. R., Carvalho, C. M., &

Scott, J. G. (2019). Targeted Smooth Bayesian Causal Forests: An

analysis of heterogeneous treatment effects for simultaneous versus


interval medical abortion regimens over gestation. arXiv preprint arXiv:1905.09405.

§ Starling, J. E., Murray, J. S., Carvalho, C. M., Bukowski, R., & Scott, J. G.

(2018). Functional response regression with funBART: an analysis of

patient-specific stillbirth risk. arXiv preprint arXiv:1805.07656.

§ Lohr, P. A., Starling, J. E., Scott, J. G., & Aiken, A. R. (2018).

Simultaneous compared with interval medical abortion regimens where

home use is restricted. Obstetrics & Gynecology, 131(4), 635-641.


J. DEVELOPMENT

SDS has partnered with the College of Natural Sciences (CNS) External Relations team to

create both short- and long-term plans for building philanthropic and community partnerships

with alumni, friends, corporations, and foundations. SDS is continuing to actively pursue

opportunities for support of programs and conferences through NIH, NSF, and the Institute of

Mathematical Statistics Meeting sponsorship.


Appendix A: SDS Organizational Chart

DeanCollege of Natural Sciences

Paul Goldbart, PhD

Chair, interimPeter Müller, PhD

SDS Core Faculty (50% Time)

Peter Müller, PhD Mathematics

Stephen Walker, PhD Mathematics

Carlos Carvalho, PhD IROM

James Scott, PhD IROM

Sinead Williamson, PhD

IROM

Cory Zigler, PhD DMS

SDS Core Faculty(0% Courtesy Appt)

Paul Damien, PhDIROM

Lauren Meyers, PhDIntegrative Biology

Jared Murray, PhDIROM

Tom Sager, PhDIROM

Tom Shively, PhDIROM

Mingyuan Zhou, PhDIROM

SDS Full Faculty

(100% Time)

Abhra Sarkar, PhD

Purnamitra Sarkar, PhD

Assistant Director for AdminstrationVicki L. Keller, MA

Coordinates daily SDS operations & student programsSupervise departmental staffGraduate Coordinator Budgets & EOMCommunicationsReportingFaculty Affairs

Adminstrative ManagerSasha Schellenberg, B.A.

Supervise classified staffCourse scheduling & catalog managementCIS, ALEKS, FASET, Final Exam scheduling, HB 2504 adminCoordinates SDS Seminar Series & Summer Statistics InstituteEvent planning

Grants & Contracts SpecialistStephanie Tomlinson

Grants: pre- and post-award processingT-32 Grant Management

Senior Administrative AssociateVictoria Obioma

HR processingPurchasingPayroll AccountingInventoryRecords MaintenanceBillingDEFINE Vouchers

Student AssistantJuan Pablo Avila

Laura Garcia (ULN)

Director of Professional Education & Consulting

Michael J. Mahometa, PhD Manages all consulting operationsOversees FAC 101B & Stat App ServerProfessional education program developmentOn-ramps & MOOC development

Staff ConsultantVACANT

Manages short coursesOn-ramps development

Staff Consultant (p/t)Erika Hale, M.S.Evan Stein, PhD

Director of Undergraduate Studies

Kristin Harvey, PhDCoordinates SDS 302Teaches undergraduate coursesCourse & Curriculum representativeManage undergraduate courses and course development

Assistant Professor of Instruction

Bindu Viswanathan, PhD

Manages Corporate Partnership programNTT course assignments & staffingTeaches graduate & undergraduate courses

Instructional Faculty

Lauren Blondeau, PhDSarah Collins, PhDSteven Hernandez, MSMatt Hersh, PhDKyongJoo Hong, PhDVera Ioudina, PhDMaggie Myers, PhDMary Parker, PhDSally Ragsdale, MSJason Rarick, MALindsey Smith, PhDNathaniel Woodward, PhD


Appendix B: SDS Core Faculty Carlos Carvalho, Professor, IROM

Dr. Carvalho’s research focuses on Bayesian statistics in high-dimensional problems with applications ranging from finance to genetics. Some of his current projects include work on causal inference, machine learning, policy evaluation and empirical asset pricing. He is also the Executive Director of the Center for Enterprise and Policy Analytics , a unit dedicate to support research, education, and dialogue around the impact of economic policies on markets and the free enterprise system. Dr. Carvalho did his undergraduate studies in Brazil. He received his

PhD in Statistics from Duke University in 2006. Before moving to The University of Texas, he was an Assistant Professor at The University of Chicago Booth School of Business. Honors:

• CBA Foundation Advisory Council Centennial Fellow—The University of Texas, Austin, 2012–

• Donald D. Harrington Faculty Fellow—The University of Texas, Austin, 2009–10 • IBM Corporation Scholar—The University of Chicago, 2008–09 • Dennis V. Lindley Prize for innovative research in Bayesian Statistics. Honorable

Mention for “Dynamic Matrix-Variate Graphical Models,” 2007 • Leonard J. Savage Award for outstanding doctoral dissertation in Bayesian

econometrics and statistics—Honorable Mention, 2006 Paul Damien, Professor, IROM

Dr. Damien’s research interests lie at the intersection of mathematics, Bayesian statistics, and applications in business, engineering, and stochastic optimization. He is a Fellow of the Royal Statistical Society of England. He is twice recipient of the United Kingdom Engineering and Physical Sciences Research Council Research Award, and was awarded the United Kingdom Overseas Visiting Fellowship by the Royal Statistical Society of England. Damien has been a Visiting Scholar at the Department of Economics in the University of Groningen, Netherlands. Dr. Damien consults for several companies in a variety of industries world-wide.

Dr. Damien received his PhD from Imperial College, UK. Honors:

• Fellow—Royal Statistical Society of England • McCombs’s School of Business Dean’s Research Fellow


Lauren Ancel Meyers, Professor, Integrative Biology

Dr. Meyers received her B.A. degree in Mathematics and Philosophy from Harvard University in 1996 and her PhD from the department of Biological Sciences at Stanford University in 2000. She joined the faculty at the University of Texas at Austin in 2003 where she was recently promoted to Full Professor and awarded a Donald D. Harrington Faculty Fellowship. She has also been an active member of the external faculty of the Santa Fe Institute since 2003 and now serves on its Scientific Advisory Board. Lauren has developed new mathematical methods for forecasting and optimizing the control of infectious diseases including meningitis, HIV, influenza, walking

pneumonia, and SARS. Her research has been published in over 45 peer-reviewed publications and funded by research grants from National Institutes of Health, the National Science Foundation, and the James S. McDonnell Foundation. The Wall Street Journal, Newsweek, the BBC, and other news sources have highlighted Lauren's work; and a number of government agencies have sought her expertise, including the Centers for Disease Control and Prevention (CDC), the Biomedical Advanced Research and Development Authority (BARDA), and the US National Intelligence Council. In 2004, the MIT Technology Review named Lauren as one of the top 100 global innovators under age 35. Honors:

• William H. and Gladys G. Reeder Faculty Fellow—The University of Texas, Austin, 2011-2013

• Donald D. Harrington Faculty Fellowship—The University of Texas, Austin, 2010-2011

• Fellow, University of Texas Institute for Molecular and Cellular Biology, 2006-2010

• College of Natural Sciences Teaching Excellence Award—The University of Texas, 2005

• MIT Technology Review TR100: One of 100 Top Global Innovators Under 35, 2004

Jared Murray, Assistant Professor, IROM

Dr. Murray’s current research interests are in developing flexible Bayesian multivariate models for heterogenous and structured data, with applications to multiple imputation for missing data, latent variable modeling and causal inference. He was recently awarded an NSF grant “Improving Probabilistic Record Linkage and Subsequent Inference” to develop new methods for matching records across files in the absence of unique identifiers, and for making inference using

the combined files.


Peter Müller, Professor, Mathematics

Dr. Müller’s interest is in methods and applications of Bayesian inference. More specifically, he is working on nonparametric Bayesian inference, decision problems, and applications to biomedical research problems. Nonparametric Bayesian inference refers to prior models for infinite dimensional random quantities, typically random probability measures. Decision problems include particular clinical trial design and multiple comparison procedures. Other applications that interest Dr. Müller include inference related to dependence structure, specifically graphical models to formalize inference about dependence for high throughput genomic data.

Another large area of application is population pharmacokinetic and pharmacodynamic models, which give rise to many good applications that exploit many of his methodological interests. Dr. Müller’s undergraduate education is from Universität Wien and Technische Universität Wien, Austria. He received his PhD from Purdue University where he worked under Jim Berger on MCMC for constrained parameter problems. He spent several years at the Institute of Statistics and Decision Sciences (ISDS), Duke University, and at M.D. Anderson Biostatistics. Honors:

• Zellner Medal, ISBA, 2019 • Fellow of the American Statistical Association • President of the International Society for Bayesian Analysis, 2010 • Robert R. Herring Distinguished Professorship in Clinical Research, 2007–2011

Tom Sager, Professor, IROM

The major part of Dr. Sager’s research focuses on the development and application of econometric models to study the empirical behavior of insurance companies. The models involve simultaneously interacting manifest and latent variables interacting in autocorrelated nonlinear panel data structures. He has applied these to the prediction of insurer insolvency, the risks of mortgage-backed securities, the effects of the Affordable Care Act, the assessment of global systemic risk, consequences of regulation, and other issues. Dr. Sager seeks to illuminate how agency theory, transactions cost economics, theories of limited and unlimited risk, and other ideas

play out to explain insurers’ management of their enterprise risks. Honors:

• Associate Editor, Journal of Risk and Insurance, 2001–2018

• Joe D. Beasley Award for teaching excellence

Department of Statistics & Data Sciences 28

Abhra Sarkar, Assistant Professor

Dr. Sarkar’s interests center around the development of novel statistical approaches that improve results and practice in an initial motivating application area while also having much broader general utility. In particular, he enjoys developing Bayesian non- and semi-parametric methods that accommodate a wide range of data generating processes, adapting to different level of data complexity and potentially automating various aspects of the analysis, including feature extraction, selection of variables, quantification of model

uncertainty and test hypothesis of interest. Dr. Sarkar earned his PhD in Statistics from Texas A&M University, and was a postdoctoral associate at Duke University. Honors:

• International Indian Statistical Association (IISA) Student Paper Award (2014)

Purnamrita Sarkar, Assistant Professor

Dr. Sarkar works on large-scale statistical machine learning problems with a focus on statistical models, asymptotic theory and scalable inference algorithms for large networks. Dr. Sarkar graduated from the School of Computer Science at Carnegie Mellon University in 2010. After earning her doctorate she was a postdoctoral scholar at U. C. Berkeley jointly in the Department of Electrical Engineering and Computer Sciences and the Department of Statistics. She received her Bachelor's degree in

Computer Science from the Indian Institute of Technology, Kharagpur in 2004. Honors:

• Best paper award, 29th International Conference on Data Engineering (ICDE), 2013

• Best paper award, International Conference on Social Networks Analysis and Mining (ASONAM), 2009


James G. Scott, Associate Professor, IROM

Dr. Scott's research focuses on statistical methodology for high-dimensional data sets, with applications in a diverse set of areas spanning the social, physical, and biomedical sciences. Three areas of methodological focus include (1) large-scale multiple testing, anomaly-detection and screening problems, where the rate of false discoveries must be controlled in order to yield viable inferences; (2) inference in sparse models; and (3) the application of data-augmentation theory and algorithms to improve the efficiency of Bayesian inference in large-scale models for discrete data sets. His recent applied work has included collaborations in health care,

demography, linguistics, biology, and neuroscience. Dr. Scott received his PhD is from Duke University, where he studied Bayesian model selection under Jim Berger. Before that he studied at Trinity College, Cambridge for two years. He was an undergraduate from 2000 to 2004 at UT-Austin in the Dean's Scholars and Plan II honors programs. Honors:

• Regents’ Outstanding Teaching Award, 2014 • NSF CAREER Grant, 2013 • Savage Award, 2010 (awarded by the International Society of Bayesian Statistics

for best thesis in Bayesian statistical theory) • National Science Foundation Graduate Research Fellowship, 2006–2009 • Marshall Scholarship for study in Great Britain, 2004–2006

Tom Shively, Professor, IROM

Dr. Shively’s research focuses on nonparametric function estimation and its application in energy economics, finance, marketing and transportation science. Current problems of interest include estimating univariate functions nonparametrically subject to shape constraints such as monotonicity and convexity, developing tests for monotonicity, and extending these results to multivariate functions using Bayesian additive regression trees. He is also interested in MCMC-based computational methods that allow for nonparametric function estimation in a wide range of non-Gaussian models including generalized mixed models, hazard function models and

models for extreme values. His recent applied research focuses on modeling electricity prices in the deregulated Australian market using multivariate nonparametric methods.

Dr. Shively received a B.A. degree in Mathematics from Middlebury College in 1981, an M.B.A. from the University of Chicago in 1984, and a PhD in Statistics from Chicago in 1986. He has been at The University of Texas at Austin since 1986.

Honors: • Outstanding Professor in the Executive MBA Program Award, 2007, 2003

• Joe D. Beasley Award for Teaching Excellence, 2001, 1993


Stephen G. Walker, Professor, Mathematics Dr. Walker’s main research focus is on Bayesian parametric and nonparametric methods. He has worked on applications, methodology, theory, implementation via MCMC, and foundational issues. Dr. Walker’s main areas of applications include medical statistics and financial data. Recent work on Bayesian nonparametrics includes constructing time series and regression models. Recent work also includes working with Bayesian models under misspecification and using loss

functions as an alternative to probability models within a learning process akin to Bayesian updating. Dr. Walker received his BA (Hons.) in Mathematics at the Oriel College of Oxford University, being awarded Open Exhibition on entry to the college. He received his PhD in Statistics from the Imperial College of London in 1995, supervised by Jon Wakefield. Dr. Walker has taught at various institutions: Imperial College at London, the University of Bath, and most recently at the University of Kent.

Honors: • Editor of Journal of Statistical Planning and Inference, 2017– • Chair of Bayesian Nonparametric Section of ISBA, 2010–2012 • EPSRC Advances Research Fellow, 2001–2006

Sinead Williamson, Assistant Professor, IROM

Dr. Williamson’s main research focus is the development of nonparametric Bayesian methods for machine learning applications. In particular, she is interested in constructing distributions over correlated measures and structures, in order to model correlated data sets or data with spatio-temporal dependence. Examples include models for documents whose topical composition varies through time, and models for temporally evolving social networks. A key research goal is the development of efficient inference algorithms for such models, and she is currently investigating methods that allow us to apply Bayesian nonparametric techniques to large datasets.

Dr. Williamson received her MEng from the University of Oxford, MSc from University College London, and PhD from the University of Cambridge. Before joining the faculty at UT Austin, she was a postdoctoral scholar at Carnegie Mellon University.


Mingyuan Zhou, Assistant Professor, IROM Dr. Zhou’s research lies at the intersection of Bayesian statistics and machine learning. He is interested in developing statistical theory and methods, hierarchical models, and efficient Bayesian inference for big data. He is currently focused on the development of nonparametric Bayesian priors for both count and mixture modeling and dictionary learning. Dr. Zhou is building the negative binomial process family to introduce new exchangeable random partitions and novel clustering algorithms. Dr. Zhou received his PhD in Electrical and Computer Engineering from

Duke University in 2013, Master's in Signal and Information Processing from the Chinese Academy of Sciences in 2008, and B.Sc. in Acoustics from Nanjing University in 2005. Honors

• McCombs Research Excellence Grant, November 2018 • CBA Foundations Research Excellence Award for Assistant Professors, April 2018 • NSF@ISBA junior travel support, ISBA World Meeting 2016 • Best Student Poster Award in NIPS 2015 Workshop: Networks

Corwin Zigler, Associate Professor, DMS

Dr. Zigler’s main research focuses on statistical methodology to confront the challenges of complex observational studies. Areas of focus include Bayesian methods, causal inference, comparative effectiveness research, spatial statistics, and environmental health data science. He has established a national and international reputation in the fields of statistics and biostatistics, and has impacted other areas of public health research including air pollution regulatory policy, environmental engineering, cancer health services research, interventional cardiology, and health care policy.

Dr. Zigler received his PhD from the Department of Biostatistics at the UCLA Fielding School of Public Health. Before joining the faculty at UT Austin, he was at the Harvard School of Public Health.

Honors • Rothman Epidemiology Prize, 2019 • American Statistical Association Statistics in Epidemiology Section Young Investigator

Award, 2012


Appendix C: SDS Seminar Series Speakers Date Speaker Name Institution Title of Talk

8/31/18 Katherine Heller Duke University “Machine Learning for Health Care”

9/21/18 Su Chen University of Texas at Austin

“Fast Bayesian Variable Selection: Solo Spike and Slab”

9/28/18 Lucas Janson Harvard University “Using Knockoffs to find important variables with statistical guarantees”

10/5/18 Carlos Pagani Zanini

University of Texas at Austin

“A Bayesian Random Partition Model for Sequential Refinement and Coagulation”

10/12/18 Mingzhang Yin University of Texas at Austin

“ARM: Augment-REINFORCE-merge gradient for discrete latent variables”

10/19/18 Debdeep Pati Texas A&M University

“Constrained Gaussian processes and the proton puzzle problem”

10/26/18 Yuguo Chen University of Illinois at Urbana- Champaign

“Statistical Inference on Dynamic Networks”

11/2/18 Evan Ott University of Texas at Austin

“Bayesian Deep Learning: Extending Probabilistic Backpropagation and Transfer Learning”

11/9/18 Mengie Wang University of Texas at Austin

“A Data Dependent Posterior Density Generative Model”

11/16/18 David Yeager University of Texas at Austin

“Heterogeneous Effects of a Scalable Growth-Mindset Intervention on Adolescents’ Educational Trajectories”

11/29/18 Sonia Petrone Universitá Bocconi “Quasi-Bayes properties of a sequential procedures for mixtures.”

11/30/18 Choudur Lakshminarayan

University of Texas at Austin

“Data Compression, and Statistical Pattern Recognition in Healthcare Applications”

12/7/18 Cory Zigler University of Texas at Austin

“Bipartite Causal Inference with Interference: Estimating Health Impacts of Power Plant Regulations”

1/23/19 Andee Kaplan Duke University “Life After Record Linkage: Tackling the Downstream Task with Error Propagation”

1/25/19 Eric Jones University of California – Berkeley

“Exploiting computational scale for richer model-based inference”

1/28/19 Maricela Cruz University of California – Irvine

“Interrupted Time Series Models for Analyzing Complex Healthcare Interventions”

1/30/19 Prasad Patil Harvard University “Replicability of genomic signatures and scientific results”

2/1/19 Rachel Nethery Harvard University “Bayesian and Machine Learning Approaches to Estimate Causal Effects in Environmental Health Applications”

2/4/19 Antonio Linero Florida State University

“Theory and Practice for Bayesian Regression Tree Ensembles”


2/22/19 Lorenzo Trippa Harvard University “Bayesian Designs for Glioblastoma Clinical Trials”

3/1/19 Raquel Prado University of California - Santa Cruz)

“Bayesian models for complex-valued fMRI”

3/8/19 Veera Baladandayuthapani

University of Michigan

“Bayesian Models for Richly Structured Data in Biomedicine”

3/11/19 Roger Peng Johns Hopkins Bloomberg School of Public Health

“Estimating the health impacts of modifying air pollution mixtures”

3/29/19 Matteo Vestrucci University of Texas at Austin

“Cognitive disease progression models for clinical trials in autosomal-dominant Alzheimer's disease”

4/5/19 Paul Rathouz University of Texas at Austin

“Semiparametric Generalized Linear Models: Small, Large, and Biased Samples”

4/26/19 Junming Yin University of Arizona

“Towards Better Learning from Crowd Labeling”

5/3/19 Surya Tokday Duke University “Quantile Regression for Correlated Response”

5/7/19 Subhajit Dutta IIT Kanpur “On Perfect Classification and Clustering for Gaussian Processes”

5/10/19 Georgia Papadogeorgou

Duke University “Unmeasured spatial confounding in air pollution studies”

Documents

Annual Report 2018-2019 · Bayesian nonparametric framework for high dimensional problems. Visiting Researchers The Department of Statistics and Data Sciences hosted five visiting