Daniel A. Spielman

I am a Sterling Professor of Computer Science and a Professor Statistics and Data Science and of Mathematics at Yale University.

I am the inagural James A. Attwood Director of the Institute for the Foundations of Data Science (FDS) and a member of TILOS, the NSF Institute for Learning-Enabled Optimization at Scale.

I was a co-Director of the Yale Institute for Network Science (YINS), a Simons Investigator and a MacArthur Fellow.

I am a member of the ACM, AMS, IEEE, SIAM, the National Academy of Sciences, American Academy of Arts and Sciences, and the CT Academy of Science and Engineering.

Descriptions of my research

The following are some articles and citations that describe important parts of my research.

The Computer Scientist Who Parlays Failures Into Breakthroughs, by Mordechai Rorvig, in June 2022 issue of Quanta Magazine.
Infinite complexity, by Richard Panek, from May/June 2022 issue of Yale Alumni Magazine.
'Outsiders' Crack 50-Year-Old Math Problem, an article by Erica Klerreich from Quanta Magazine about the solution of the Kadison-Singer Problem.
Network Solutions, an article by Erica Klarreich for the Simons Foundation about work in which my community is involved.
Gil Kalai's laudation for my Nevanlinnna Prize.
An article about by research by Michel Goemans and Jonathan Kelner.
An article for SIAM News by Dana Mackenzie about my research on the Kadison-Singer Problem.
The citation for the 2008 Godel Prize.
The citation for the 2015 Godel Prize.
The article "Smoothed Move" by Barry Cipra in What's Happening in the Mathematical Sciences, Volume 6, an article about Smoothed Analysis.
The article "The Kadison-Singer Problem: A Fine Balance" by Dana Mackenzie in What's Happening in the Mathematical Sciences, Volume 10.

Slides from and videos of some talks

Spectral Graph Theory and its Applications, a tutorial I gave at FOCS 2007.
My talk from ICM 2010: slides, video, paper, opening ceremony.
The Blyth Memorial Lectures at Toronto on Laplacian Matrices of Graphs: Applications (9/28/11), Computations (9/29/11), and Approximations (9/30/11).
Spectral and Electrical Graph Theory (given at the Caesarea Rothschild Institute, Haifa, May 17, 2011.
Spectral Sparsification of Graphs (as given at the Weizmann Institute on May 15, 2011). A video of me giving a related talk at MSR NE
FOCS 2010
EPFL Sparsification Talk, from the June 2012 Algorithmic Frontiers Workshop.
The Erdos Lectures at Hebrew University (2014). 1. Kadison-Singer, 2. Sparsification of Graphs and Matrices, 3. Ramanujan Graphs of Every Degree.
Slides from my Hamilton Day Lecture, in Powerpoint (recommended) or PDF.
Slides from the 2016 Josiah Willard Gibbs Lecture at the Joint Mathematics Meeting, in Powerpoint (recommended) or PDF.
Slides from my plenary talk at ICML 2016 in Powerpoint (recommended) or PDF.
Slides from my plenary talk at ISIT 2016 in Powerpoint (recommended) or PDF.
The video from my AMS/MAA invited address at the 2019 Joint Mathematics Meetings.
The video of my 1W-MINDS talk on Balancing Covariates in Randomized Experiments.
The video of my lecture "Discrepancy Theory and Randomized Controlled Trials" given at the Harvard Math-Science Literature Lecture Series.
My talk on balancing covariates from Foundations of Computational Mathematics, 2020, rescheduled for 2021. (the first hour is my talk)

My Courses

Computation and Optimization (Spring 2020).

Spectral Graph Theory (Fall 2019),

Software by my group

Discrepancy Minimizer An app to partition rows of spreadsheets into groups of low discrepancy. By Deyuan Li.
Laplacians.jl a Julia package for solving problems related to Laplacian matrices and for generating interesting problem instances. It contains good code for solving Laplacian linear equations, computing low-stretch spanning trees, and generating interesting graphs (the Chimera generator).
GSWDesign An implementation of the Gram-Schmidt Walk Design described in the paper
IJuliaTimeMachine An extention of IJulia (for Jupyter notebooks) that allows one to retrieve variables from old cells and to run cells in threads.

Balancing covariates in randomized experiments using the Gram-Schmidt walk

YINSlex The code used in the experiments from the paper Algorithms for Lipschitz Learning on Graphs
Isotonic An interior point method for isotonic regression. Based on the paper Fast, Provable Algorithms for Isotonic Regression in all l_p-norms. Laplacians.jl contains a variant of this code.