DIANA CAI

University of Chicago, Department of Statistics
Email: dcai-at-uchicago-dot-edu

Research interests: I am broadly interested in statistical modeling, theory, and applications, with an emphasis on scalable, nonparametric learning.

Recent News

Selected Work

View all papers

About Me

I am a Ph.D. student in Department of Statistics at the University of Chicago. I received an A.B. in computer science and statistics from Harvard University. I was a member of the Harvard Intelligence Probabilistic Systems (HIPS) Group, and was advised by Ryan P. Adams. I have also worked on applying machine learning to natural language processing problems at several startups and am excited about developing new machine learning methodology that can be applied in practice.

Projects

Projects, by topic:

  1. graphs and networks,
  2. time series modeling,
  3. other projects.


Research Projects



Graphs and networks



Edge-exchangeable graphs, sparsity, and power laws

(with Trevor Campbell and Tamara Broderick).

Advances in Neural Information Processing Systems (NIPS), to appear, 2016. Completely random measures for modeling power laws in sparse graphs. NIPS workshop on Networks in the Social and Informational Sciences, 2015. [[pdf]]
Edge-exchangeable graphs and sparsity. NIPS workshop on Networks in the Social and Informational Sciences, 2015. [[pdf]]
Edge-exchangeable graphs, sparsity, and power laws. NIPS Workshop on Bayesian Nonparametrics: The Next Generation, 2015. [[pdf]]

Priors on exchangeable directed graphs

(with Nate Ackerman and Cameron Freer).

Exchangeable directed graphs are characterized by a sampling procedure given by the Aldous-Hoover theorem, determined by specifying a distribution on measurable objects known as digraphons. We present a new Bayesian nonparametric model for exchangeable directed random graphs.
Electronic Journal of Statistics (EJS), to appear, 2016.
In the NIPS Workshop on Bayesian Nonparametrics, 2015. [ arXiv ]
10th Conference on Bayesian Nonparametrics, 2015.

Iterative step-function estimation for graphons

(with Nate Ackerman and Cameron Freer).

A method for estimating graphons (symmetric, measurable functions from which we can sample exchangeable graphs) by iteratively forming step-functions obtained by grouping vertices by edge density.
Submitted, 2016. [ PDF | arXiv ]
In the Women in Machine Learning Workshop, 2014.



Times series modeling



Efficient Variational Approximations for
Online Bayesian Changepoint Detection

(with Ryan P. Adams).

We develop a scalable method for online changepoint detection in latent variable models using sparse variational approximations.
[ Github ]
New England Machine Learning Day Workshop, 2014.
In the Women in Machine Learning Workshop, 2013.



Other projects



The Ratio Project: Analyzing Online Recipes

We analyzed online recipes using computational methods and exploratory visualizations.
The Boston Globe, Dec 2011. [[link]]
Bakery Mapping Visualization: [[pdf]] | Slides: [[pdf]]
(with Elaine Angelino and Michael Brenner)
Cocktails Visualization, Jun 2013.
(with Elaine Angelino, Gabrielle Ehrlich, Brent Heeringa, Michael Mitzenmacher, Naveen Sinha).


Papers


Working Papers

  1. An iterative step-function estimator for graphons.
    Diana Cai, Nate Ackerman, Cameron Freer.
    arXiv:1412.2129 [math.ST, stat.ML, stat.CO].

Journal and Conference Papers

  1. Edge-exchangeable graphs and sparsity.
    Diana Cai, Trevor Campbell, Tamara Broderick.
    Advances in Neural Information Processing Systems (NIPS), to appear, 2016.
  2. Priors on exchangeable directed graphs. arxiv:1510.08440.
    Diana Cai, Nate Ackerman, Cameron Freer.
    Electronic Journal of Statistics (EJS), to appear, 2016.

Workshop Papers

  1. Priors on exchangeable directed graphs.
    Diana Cai, Nate Ackerman, Cameron Freer.
    NIPS Workshop on Bayesian Nonparametrics: The Next Generation, 2015. [[pdf]]
    ISBA@NIPS Special Travel Award for Contributed Paper, 2015.
  2. Completely random measures for modeling power laws in sparse graphs.
    Diana Cai, Tamara Broderick.
    In the NIPS workshop on Networks in the Social and Informational Sciences, 2015. [[pdf]]
    arxiv:1603.06915 [stat.ML, math.ST, stat.ME].
  3. Edge-exchangeable graphs, sparsity, and power laws.
    Tamara Broderick, Diana Cai.
    In the NIPS Workshop on Bayesian Nonparametrics: The Next Generation [[pdf]]
    ISBA@NIPS Special Travel Award for Contributed Paper, 2015.
    In the NIPS workshop on Networks in the Social and Informational Sciences, 2015. [[pdf]]
    arxiv:1603.06898 [math.ST, stat.ME, stat.ML].

Theses

  1. Scalable methods for Bayesian online changepoint detection.
    Advisor: Ryan P. Adams
    Senior Thesis, Harvard University, 2014.


Selected Talks

  1. Edge-exchangeable graphs, sparsity, and power laws.
    Invited talk in the Isaac Newton Institute (INI) workshop on Bayesian methods for networks, July 2016. [[video link]]
  2. Edge-exchangeable graphs, sparsity, and power laws.
    Invited talk at the Massachusetts Institute of Technology, Machine Learning Tea seminar, July 2016.
  3. Edge-exchangeable graphs, sparsity, and power laws.
    Contributed talk in the NIPS workshop on Bayesian Nonparametrics: the Next Generation, December 2015. [pdf]
  4. Priors on exchangeable directed graphs.
    Contributed talk in The 10th Conference on Bayesian Nonparametrics, June 2015. [pdf]
  5. Efficient online variational changepoint detection.
    Machine Learning Tea Seminar, Harvard University, Feb 2013.

Poster presentations

  1. Completely random measures for modeling power laws in sparse graphs.
    NIPS workshop on Networks in the Social and Informational Sciences, Dec 2015.
  2. Edge-exchangeable graphs, sparsity, and power laws.
    NIPS workshop on Bayesian Nonparametrics: The Next Generation , Dec 2015.
    NIPS workshop on Networks in the Social and Informational Sciences, Dec 2015.
  3. Priors on exchangeable directed graphs.
    NIPS workshop on Bayesian Nonparametrics: The Next Generation, Dec 2015.
    Women in Machine Learning Workshop, Dec 2015.
  4. An iterative step-function estimator for graphons.
    Women in Machine Learning Workshop, Dec 2014.
  5. Efficient variational approximations for online Bayesian changepoint detection. New England Machine Learning Day Workshop, May 2014.
  6. Efficient variational approximations for online Bayesian changepoint detection. Women in Machine Learning Workshop, Dec 2013.

Teaching

University of Chicago

  • STAT 220000: Statistical Methods and Applications. Teaching Assistant: Winter 2016, Spring 2016

Harvard University

  • CS181: Machine Learning. Teaching Fellow, Spring 2014