PeopleofSuperintelligence

Researcher

Michal Derezinski

Also known as Michał Dereziński

Michal Derezinski is a computer scientist whose research on randomized algorithms for large matrices and data sampling serves machine learning and data science. After master's degrees at the University of Warsaw, he earned a PhD in computer science at UC Santa Cruz in 2018 under Manfred Warmuth and later held a postdoctoral fellowship in statistics at UC Berkeley. In 2020, a paper with Rajiv Khanna and Michael Mahoney on column subset selection and the Nyström method won a NeurIPS Best Paper Award. He joined the University of Michigan in 2021, and in 2024 he received an NSF CAREER Award for a project on randomized linear algebra for data science. He is an assistant professor of computer science and engineering at Michigan, where he works on randomized algorithms for machine learning, optimization and applied mathematics.

Timeline

  1. 2017

    Volume sampling for regression

    With his doctoral advisor Manfred Warmuth, he posts "Unbiased estimates for linear regression via volume sampling" in 2017. The paper shows that choosing columns in proportion to the squared volume they span makes the least-squares solution from a small labeled subset an unbiased estimator of the solution on all labels.

  2. 2018

    PhD and Simons Institute fellowship

    After master's degrees in mathematics and computer science at the University of Warsaw, he completes a PhD in computer science at UC Santa Cruz in 2018, and his work on sampling methods in statistical learning earns the Best Dissertation Award. In Fall 2018, he is a Patrick J. McGovern Research Fellow in the Foundations of Data Science program at the Simons Institute for the Theory of Computing.

  3. 2019

    Fast determinantal point processes

    In 2019, he publishes "Fast determinantal point processes via distortion-free intermediate sampling" at the Conference on Learning Theory, a sampling algorithm whose sampling step takes time independent of the number of rows. Determinantal point processes are used in machine learning, stochastic optimization, data summarization and low-rank matrix reconstruction.

  4. 2020

    A NeurIPS Best Paper Award

    In 2020, he, Rajiv Khanna and Michael W. Mahoney, all at UC Berkeley, win a NeurIPS Best Paper Award for "Improved guarantees and a multiple-descent curve for Column Subset Selection and the Nystrom method". The paper gives sharper approximation guarantees for these low-rank approximation tools and shows that the approximation factor can rise and fall several times as the subset size k grows.

  5. 2021

    Joining the Michigan faculty

    He joins the University of Michigan's Computer Science and Engineering faculty as an assistant professor for the 2021/22 academic year, coming from a postdoctoral fellowship in the Department of Statistics at UC Berkeley.

  6. 2023–2024

    A tutorial on randomized linear algebra

    With Michael W. Mahoney, he presents the NeurIPS 2023 tutorial "Recent and Upcoming Developments in Randomized Numerical Linear Algebra for ML". In 2024, the two post a written overview of the field, which covers randomized algorithms for the large matrices that hold datasets, graphs and model weights.

  7. 2024

    An NSF CAREER Award

    In 2024, he receives a National Science Foundation CAREER Award for the project "Leveraging Randomization and Structure in Computational Linear Algebra for Data Science". The project aims to design and deploy improved linear algebra algorithms for computational data science.

  8. As of October 2026

    Assistant professor at Michigan

    He is an assistant professor of computer science and engineering at the University of Michigan, where his research addresses the theoretical foundations of randomized algorithms for machine learning, optimization and applied mathematics. In Fall 2025, he is a visiting scientist in the Simons Institute's Complexity and Linear Algebra program, and in October 2025 he gives the Stanford ISL Colloquium talk "Randomized Algorithms for Solving Linear Systems with Low-Rank Structure".

Career

  1. Since 2021University of MichiganAssistant Professor
  2. Until 2021University of California, BerkeleyPostdoctoral Fellow
  3. 2018Simons Institute for the Theory of ComputingPatrick J. McGovern Research Fellow, Visiting Scientist

Education

  1. 2018University of California, Santa CruzPhD
  2. University of WarsawMaster's degrees in mathematics and computer science

From the biography.

  1. The Massachusetts Institute of Technology trained or employed 44 of the 187 AI major-prize winners, more than any other universityIncludes Michal Derezinski. The Massachusetts Institute of Technology trained or employed 44 of the 187 AI major-prize winners, more than any other university. We compare doctorate and 5+ year employment links, career timing and current workplaces for 187 major-prize and 1,111 best-paper winners.
  2. Universities had a net outflow of 2,068 moves in AI careersIncludes University of Michigan. Our estimates of career flows and median h-index across universities and companies in AI, including the role of joint appointments and earlier academic jobs.
  3. Universities keep AI researchers for a median of 6 years, compared with 2 years at frontier labsIncludes University of Michigan. Across 15,811 dated work stays (6,529 people), median role tenure runs from 6 years at universities to 2 years at frontier labs. We compare all five organization types, starting decades from the 1970s to the 2020s, and the organizations that keep people longest.
  4. The median h-index is 48 at universities and 14 at frontier AI labsIncludes University of Michigan. The median h-index is 48 at universities and 14 at frontier AI labs. We compare publishing careers, recent citations, dual appointments and job moves across 2,420 university researchers and 308 frontier-lab researchers.

Colleagues

People named in this biography, and people with dated positions at the same organization in the same years.

Sources and credits

Photo: We have no published photo of Michal Derezinski yet. Send us a free-licensed photo.

  1. Michał Dereziński web.eecs.umich.edu. Checked 21 October 2026.
  2. Unbiased estimates for linear regression via volume sampling arxiv.org. Checked 21 October 2026.
  3. CSE welcomes new faculty for academic year 2021/22 web.archive.org. Checked 21 October 2026.
  4. Michal Derezinski simons.berkeley.edu. Checked 21 October 2026.
  5. Fast determinantal point processes via distortion-free intermediate sampling proceedings.mlr.press. Checked 21 October 2026.
  6. Announcing the NeurIPS 2020 award recipients blog.neurips.cc. Checked 21 October 2026.
  7. Improved guarantees and a multiple-descent curve for Column Subset Selection and the Nystrom method proceedings.neurips.cc. Checked 21 October 2026.
  8. Recent and Upcoming Developments in Randomized Numerical Linear Algebra for ML neurips.cc. Checked 21 October 2026.
  9. Recent and Upcoming Developments in Randomized Numerical Linear Algebra for Machine Learning arxiv.org. Checked 21 October 2026.
  10. Michał Dereziński receives NSF CAREER Award to support next generation of data science algorithms web.archive.org. Checked 21 October 2026.
  11. Michał Dereziński | ISL Colloquium isl.stanford.edu. Checked 21 October 2026.