HOME ABOUT PAPERS TEACHING NEWS CONTACT
     

Collin A. Politsch

Ph.D. in Statistics and Machine Learning

Carnegie Mellon University

About Me


I recently completed my Ph.D. at Carnegie Mellon University, jointly studying in the Department of Statistics & Data Science and the Machine Learning Department. The area of research I am most passionate about is the intersection of statistics, machine learning, and cosmology. Much of my Ph.D. work focused on optimally inverting the silhouettes of the intergalactic medium in the spectra of high redshift quasars to reconstruct one, two, and three-dimensional maps of the Universe's large-scale structure. On a more theoretical level, my interests span various aspects of supervised learning, spatio-temporal modeling, and uncertainty quantification. I previously received an M.S. in machine learning from CMU and a B.S. in mathematics from the University of Kansas.

In my free time I enjoy athletics, reading, traveling, any food wrapped in a tortilla, and my dogs -- Seth and Max.

I changed my last name from Eubanks to Politsch in March of 2019 to honor my late mother.

Publications and technical reports


  1. Trend Filtering - I. A Modern Statistical Tool for Time-Domain Astronomy and Astronomical Spectroscopy [Link] [arXiv] [GitHub]
    Collin A. Politsch, Jessi Cisewski-Kehe, Rupert A.C. Croft, and Larry Wasserman
    Monthly Notices of the Royal Astronomical Society, 492(3), p. 4005-4018, 2020, DOI: 10.1093/mnras/staa106

  2. Finalist for best paper in the ASA Astrostatistics Student Paper Competition, sponsored by the Astrostatistics Interest Group. Awarded at JSM in August 2020.


  3. Trend Filtering - II. Denoising Astronomical Signals with Varying Degrees of Smoothness [Link] [arXiv] [GitHub]
    Collin A. Politsch, Jessi Cisewski-Kehe, Rupert A.C. Croft, and Larry Wasserman
    Monthly Notices of the Royal Astronomical Society, 492(3), p. 4019-4032, 2020, DOI: 10.1093/mnras/staa110

  4. Finalist for best paper in the ASA Astrostatistics Student Paper Competition, sponsored by the Astrostatistics Interest Group. Awarded at JSM in August 2020.


  5. Mapping the Large-Scale Universe through Intergalactic Silhouettes [Link]
    Collin A. Politsch and Rupert A.C. Croft
    CHANCE, 32(3), 14-19, 2019, DOI: 10.1080/09332480.2019.1662696

  6. Augmenting Adjusted Plus-Minus in Soccer with FIFA Ratings [arXiv] [Website]
    Francesca Matano, Lee F. Richardson, Taylor Pospisil, Collin Eubanks, and Jining Qin
    Preprint. arXiv:1810.08032. http://intraocular.net

  7. Multi-resolution Regression, Divide and Conquer Risk Estimation, and the Large-scale Universe [Link] [Slides]
    Collin A. Politsch. Ph.D. thesis proposal (supervised by Larry Wasserman, Jessi Cisewski-Kehe, and Rupert A.C. Croft)

  8. Convex Optimization Methods for Quantile Regression [Link]
    Purvasha Chakravarti, Collin Eubanks, and Jining Qin.

Teaching


Teaching Assistant

– 10/36-702 (StatDS/ML Ph.D. course),  Statistical Machine Learning,  Head TA,  Prof: Larry Wasserman
– 10/36-705 (StatDS/ML Ph.D. course),  Intermediate Statistics,  Head TA,  Prof: Larry Wasserman
– 36-618 (StatDS M.S. course),  Experimental Design & Time Series Analysis,  Head TA,  Prof: Edward Kennedy
– 36-467/667 (StatDS B.S./M.S. course),  Special Topics: Data over Space & Time,  Head TA,  Prof: Cosma Shalizi
– 36-402/608 (StatDS B.S./M.S. course),  Advanced Methods for Data Analysis,  Prof: Cosma Shalizi
– 36-401/607 (StatDS B.S./M.S. course),  Modern Regression,  Head TA,  Prof: Larry Wasserman
– 36-226 (StatDS B.S. course),  Introduction to Statistical Inference,  Head TA,  Instructor: Amanda Luby
– 36-225 (StatDS B.S. course),  Introduction to Probability Theory,  Head TA,  Instructor: Amanda Luby
– 36-217 (StatDS B.S. course),  Probability Theory and Random Processes,  Head TA,  Prof: Sam Ventura
– 36-217 (StatDS B.S. course),  Probability Theory and Random Processes,  Head TA,  Instructor: Ross O'Connell

Advisor

Student: Benjamin Leroy (University of California, Berkeley)
Project: Dynamical Mass Measurements of Galaxy Clusters

News


01/2020 - My paper Trend Filtering: A Modern Statistical Tool for Time-Domain Astronomy and Astronomical Spectroscopy (published as a two-part series) was selected as a finalist for the best paper in the ASA Astrostatistics Student Paper Competition, sponsored by the Astrostatistics Interest Group. I will be presenting this work at JSM in August 2020.

10/2018 - Three of my classmates and I took 2nd place in The Data Open at CMU hosted by Citadel and Correlation One (300+ applications, ~100 selected to compete).

05/2018 - I will be interning as a data scientist at Uber HQ in San Francisco during the summer of 2018.

11/2017 - CMU Statistics and Data Science Graduate Students Keep Winning Big (article detailing hackathon success)

09/2017 - Three of my classmates and I took 2nd place in the 2017 NBA Hackathon (900+ applications, ~200 selected to compete). [Recap]

09/2017 - Three of my classmates and I took 2nd place in The Data Open at CMU hosted by Citadel and Correlation One (550+ applications, ~100 selected to compete).

05/2017 - I successfully defended my thesis proposal: Multi-resolution Regression, Divide and Conquer Risk Estimation, and the Large-scale Universe. [Document] [Slides]

05/2017 - Three of my classmates and I took 3rd place in the Google BrainHub Neurohackathon, hosted by the Machine Learning Department at CMU.

Contact


Email: capolitsch [at] cmu [dot] edu