Schedule

2021 Conference
Join us April. 23-24, 2021, for the Upstat Annual Conference.

Schedule



Friday, April 23 2021

3:30 - 4:00 PM: Welcome Session

The Organizing Committee of the UPSTAT 2021 conference welcomes each and all of you attending from all round the world. We thank you for joining us for this year’s conference as together we focus on the fascinating yet incredibly challenging theme “Statistics Serving Social Justice”. Read more about the 2021 theme.

We thank our Mother Organization, the American Statistical Association (ASA) for supporting the concerted efforts of this most august organizing committee.

In the interest of guaranteed a smooth and pleasant and positively memorable conference for all, I call upon all in attendance to seriously uphold the honorable etiquette of all ASA meetings found at https://www.amstat.org/ASA/Meetings/Meeting-Conduct-Policy.aspx Please be sure to read and familiarize yourself with it.

https://us02web.zoom.us/j/8565804058

4:00 - 5:00 PM: Keynote Speakers

Leveraging Criminal Justice Data to Shed Light on Racial and Other Disparities in the System

Gipsy Escobar, PhD, Director of Product Strategy (read bio)
Sema Taheri, PhD, Director of Research Operation (read bio)


Abstract: Common ground is in short supply. The same goes for accountability and transparency. As a result, public trust in the criminal justice system is at an all time low. In this climate, lasting reform is extremely difficult to achieve. Measures for Justice (MFJ) is a non-partisan, non-profit organization with a mission to arm communities and practitioners with their own data to create sustainable solutions to systemic problems in criminal justice. When the public, police, prosecutors, and courts share and trust the same data, those data become the common ground for dialog and reform. Dr. Escobar and Dr. Taheri will discuss an initiative to partner with prosecutor offices across the country to develop a public dashboard where everyone in the community can: (1) track progress toward a shared criminal justice goal; (2) track monthly data on how cases flow through the system; and (3) break down all data by demographics like race and ethnicity, sex, age, indigent status, homelessness, among other things. They will discuss the tools used to turn prosecutor case management data into actionable, measurable information; the challenges of working with criminal justice data; a model for the community and the prosecutor to work together to co-create a joint space for making criminal justice data transparent and shared goals public; and plans to integrate data from other areas of criminal justice, including policing.


https://us02web.zoom.us/j/8565804058

5:00 - 6:00 PM: Roundtable Discussion

An expert roundtable on the promises and problems of data science in the service of social justice

Moderators:

  • Dr John McCluskey, Professor, Department of Criminal Justice, Rochester Institute of Technology
  • Dr John Handley, University of Rochester
Panelists:
  • David Banks, Duke University
  • Veronica Ciocanel, Duke University
  • William Cipolli, Colgate University
  • Gipsy Escobar, Measures For Justice
  • Arindam Gosh, Office of Criminal Justice Policy (Bexar County Texas)
  • Nicholas Robertson, Rochester Institute of Technology
  • Sema Taheri, Measures For Justice

https://us02web.zoom.us/j/8565804058

Saturday, April 24 2021

9:00 - 10:00 AM: Tutorial 1

Introduction to Python and Pandas for Data Analysis
Presenter: Vi Ly

This tutorial will be a quick introduction to data analysis with Python and pandas package, using some toy datasets. Attendees of the Python Tutorial will have the option of running the code alongside the presentation. In order to do so, attendees are recommended to download Python & Pandas through Anaconda (it should be free - install the corresponding version for your operating system and use default settings): https://www.anaconda.com/products/individual The code for the tutorial can be downloaded here: https://github.com/viquangly/Training-Materials. During the day of the tutorial, please have the Python Tutorial.ipynb & mpg dataset.csv on your desktop for easy access.

https://us02web.zoom.us/j/8565804058

10:00 - 11:00 AM: Tutorial 2

20 different ways to look at a flower: an introduction to R programming language
Presenter: Gregory Babbitt

This tutorial will demonstrate 20 different R programming scripts that examine Fisher’s iris dataset in 20 different ways. Methods will range from simple descriptive statistics and hypothesis tests, all the way to machine learning implemented on big data representing a half million flowers. The basics of R programming covered in this tutorial will include package management, data input/output, subset/reshaping data, scientific graphics with ggplot2, and a variety of machine learning packages. Please bring a laptop. All 20+ R scripts are available at this website https://gbabbitt.github.io/RocASAsamples/

https://us02web.zoom.us/j/8565804058

11:00 - 12:00 PM: Tutorial 3

Using Jigsaw puzzles to teach feature selection, teamwork, and diversity
Presenter: Thomas Kinsman

Custom designed jigsaw puzzles were created to teach feature selection to students of Data Science and Computer Vision courses. A team-oriented workshop was developed and presented to students. The jigsaw puzzles cannot be easily solved without using a diverse set of features, hence the importance of having a diverse set. However, the workshop was carefully rigged to prevent any one team member from seeing all of the features, forcing participants to collaborate. Some jigsaw puzzles were printed all one color, making them difficult to solve, and emphasized the need for diverse features and viewpoints to help solve the problem. In the final stage of the workshop, the students were taught a shared solution strategy which drastically improved their solution speed. The workshop curriculum points out analogies to people from diverse backgrounds, who are deaf, and who are colorblind. Participants overwhelmingly reported that this experience helped them understand the importance of a diversity of viewpoints, the importance of combining features, and the importance of teamwork in Data Science and Computer Vision. The students enjoyed this workshop, and learned about feature selection. This workshop reports on the puzzle design, the surprising revelation of how the workshop curriculum was rigged, and the results of the original, multi-year study.

https://us02web.zoom.us/j/8565804058

12:00 - 2:00 PM: Poster Sessions

https://us02web.zoom.us/j/8565804058

2:00 PM - 2:45 PM Parallel Sessions 1

Session Organizer/Chair: Trijya Singh

  • Talk 1
    Evaluating the Distribution of Hiring and Housing Discrimination Against Black Americans. - Dr. William Cipolli, Colgate University

  • Talk 2
    The Impact of Crowded Housing on COVID-19 Transmission Dynamics in New York City - Sara Venkatraman*, Cornell University

  • Talk 3
    Ban the Box Policies in Housing and Their Influence on Racial Discrimination - Cheyanne Simpson**, University of St. Thomas (St. Paul)

https://us02web.zoom.us/j/9365155985

Session Organizer/Chair: Tony Wong

  • Talk 1
    Betacoronavirus binding dynamics relevant to the functional evolution of the highly transmissible SARS-CoV-2 variant N501Y - Dr. Gregory Alan Babbitt, RIT

  • Talk 2
    The "lab bus:" assessing the effectiveness of a mobile community COVID-19 testing facility in reaching underserved areas - Christopher W. Ryan, SUNY Upstate Medical University & Broome County Health Department

  • Talk 3
    Summary of a Year of Work in Covid-19 Statistical Modeling - Dr. Jorge L. Romeu, State University of New York

https://us02web.zoom.us/j/9210831432

Session Organizer/Chair: Ernest Fokoue

  • Talk 1
    Order-restricted Bayesian Inference and Optimal Designs for the Simple Step-stress Accelerated Life Tests under Progressive Type-I Censoring based on Three-parameter Gamma Prior - Crystal Wiedner*, University of Texas at San Antonio

  • Talk 2
    Bayesian bridge quantile regression for high-dimensional data - Shen Zhang*, University of Texas at San Antonio

  • Talk 3
    Bayesian Bradley-Terry Modeling with Multiple Game Outcomes with Applications to European and College Hockey - Jacob Klein*, RIT

https://us02web.zoom.us/j/3633738481

Session Organizer/Chair: Katherine Grzesik

  • Talk 1
     Multivariate Semiparametric Monitoring Techniques for Mixed-Type Data and Applications to a Patient-Centered Study - Elisavet M. Sofikitou, University at Buffalo

  • Talk 2
    Predicting Tax fraud using Supervised Machine Learning approach: Case Study Rwanda - Belle Fille MURORUNKWERE*, African Center of Excellence in Data Science

  • Talk 3
     Understand protein structure with statistics - Khoa Hoang**, University of Rochester

https://us02web.zoom.us/j/6086982550

* to get the Zoom link choose one of the track sessions from above

3:00 PM - 3:45 PM Parallel Sessions 2

Session Organizer/Chair: Michael McDermott

  • Talk 1
    Proving Racial Profiling in Court - Dr. Mary S. Fowler, Worcester State University

  • Talk 2
    Analyzing Federal Sentence Disparities through Inferred Sentencing Records - Dr. Maria-Veronica Ciocanel, Duke University

  • Talk 3
    Algorithmic analysis of the Pennsylvania Additive Classification Tool - Swarup Dhar**, Bucknell University

https://us02web.zoom.us/j/9365155985

Session Organizer/Chair: Tony Wong

  • Talk 1
    Analysis of the Evolution of Parametric Drivers of High-End Sea-Level Hazards - Alana Hough*, RIT

  • Talk 2
    The Price of Procrastinating Coastal Adaptation to Sea-Level Rise - Hanna Sheets*, RIT

  • Talk 3
     Learning a Statistical Manifold to Determine the Risk Assessment Strategies for Tornado-induced Property Losses in the United States - Thilini Mahanama*, Texas Tech University

https://us02web.zoom.us/j/9210831432

Session Organizer/Chair: Yan Zhuang

  • Talk 1
     Estimation of Finite Population Size by Sequential Sampling Methods - Dr. Debanjan Bhattacharjee, Utah Valley University

  • Talk 2
    Minimum Risk Point Estimation for the Mean of a Normal Population by Sampling in Groups - Dr. Zhe Wang, Denison University

  • Talk 3
    Fixed-Accuracy Confidence Interval Estimation of P(X > c) for a Two-Parameter Gamma Population - Dr. Jun Hu, Oakland University

https://us02web.zoom.us/j/3633738481

Session Organizer/Chair: Katherine Grzesik

  • Talk 1
     Combating the Curse of Dimensionality in Multidimensional Functional Data Analysis - Will Consagra*, University of Rochester

  • Talk 2
    Misunderstood Issues in the Measurement of Demographic Differences - James P. Scanlan, Attorney at Law

  • Talk 3
     A Python Environment for Numerical Continuation Methods - Jafar Anafi*, African Institure for Mathematical Science in Cameroon

https://us02web.zoom.us/j/6086982550

4:00 PM - 4:45 PM Parallel Sessions 3

Session Organizer/Chair: Ernest Fokoue

  • Talk 1
    U.S. Fatal Police Shooting: Media Bias, Washington Post Dataset Analysis, and Fatal Police Shooting Rate and Victims’ Race Prediction - Yangxin Fan*, University of Rochester

  • Talk 2
    The Relationship Between Police Funding and Fatal Police Shootings in the United States - Arbaaz Mohideen*, North Carolina A&T State University

  • Talk 3
    Exploring Racial Disparities in NY State Patrol Stop-and-Frisk Data - Kelsey Oliverio* & Alayna Riefer*, SUNY Brockport

https://us02web.zoom.us/j/9365155985

Session Organizer/Chair: Katherine Grzesik

  • Talk 1
     AI-Driven Solution for Health-Related Social Change - Dr. Srikanta Banerjee, Walden University

  • Talk 2
    Are Deep Neural Networks Predictively Better Than Kernel Learning Machines? - Eddie Pei*, RIT

  • Talk 3
     Detection of multiple undocumented change-points using Bayesian adaptive lasso with quantile regression models - Ranran Chen*, University of Texas at San Antonio

https://us02web.zoom.us/j/9210831432

Session Organizer/Chair: Trijya Singh

  • Talk 1
     Visual Insights into the Mahalanobis Distance - Dr. Thomas Kinsman, RIT

  • Talk 2
    The Role of Kernels in Data Analysis - Dr. Marianthi Markatou, State University of New York at Buffalo

  • Talk 3
     Inference and Optimal Design for Progressively Type-I Censored Step-Stress ALT with a Log-Location-Scale Distribution - Aruni Jayathilaka*, University of Texas at San Antonio

https://us02web.zoom.us/j/3633738481

Session Organizer/Chair: Ernest Fokoue

  • Talk 1
     Application of a Finite Weakest-Link Load-Sharing Model to the Failure Process of a High-K Gate Dielectrics of a Semiconductor - Dr. David Han, University of Texas at San Antonio

  • Talk 2
    Applying Blob Detection Method to Track Atoms in TEM Images of Material Science - Yuchen Xu*, Cornell University

https://us02web.zoom.us/j/6086982550

5:00 - 6:00 PM - Data Competition

How can statistics and data science be harnessed to help define, conceptualize, operationalize, contextualize and visualize disparity in policing with the finality of hopefully infusing equity and fairness and justice

20 Teams from all around the world registered to participate in this competition

The provisional results are in and the following teams will be presenting to finally declare the winners

  • UofR-STATs  [University of Rochester]
  • FunkyStats   [University of Rochester]
  • Pinta             [Rochester Institute of Technology]


https://us02web.zoom.us/j/8565804058

6:00 - 6:30 PM - Awards & Farewell

Our undergraduate and graduate student participants are extremely precious to us, being as they are, the future of our discipline. Students contributing oral presentations and poster presentations will be assessed by special judges designated by our conference Program Chair, Dr Katherine Grzesik. The most outstanding will receive awards:

  • Gold
  • Silver
  • Bronze
  • Honorable Mention


https://us02web.zoom.us/j/8565804058