Mikhail Belkin
Also known as Misha Belkin
Mikhail Belkin is a professor at UC San Diego whose work ties the classical bias-variance trade-off to modern machine learning. He co-wrote Laplacian Eigenmaps with Partha Niyogi at NIPS 2001, a widely used algorithm. He completed a mathematics PhD at the University of Chicago in 2003, advised by Niyogi. In 2019 he and three co-authors published a PNAS paper proposing a 'double descent' curve that subsumes the textbook bias-variance trade-off. The ACM named him one of its 2023 Fellows, and he is editor-in-chief of the SIAM Journal on Mathematics of Data Science.
- Now at
- University of California, San Diego and Amazon
- Works
- 134
- Citations
- 22,605
- h-index
- 36
Publication numbers from OpenAlex, as of 15 October 2026.
Timeline
-
2001
Laplacian Eigenmaps at NIPS
He and Partha Niyogi publish Laplacian Eigenmaps at NIPS 2001, an algorithm that is widely used.
-
2003
A mathematics PhD in Chicago
He completes a mathematics PhD at the University of Chicago in 2003, advised by Partha Niyogi.
-
2019
A double descent curve
He and three co-authors publish a PNAS paper in 2019 that reconciles modern machine-learning practice with the classical bias-variance trade-off. Their 'double descent' curve subsumes the textbook U-shaped trade-off curve.
-
2023
An ACM Fellowship
The ACM names him one of its 2023 Fellows, a class chosen by peers for contributions to computing science and technology.
-
As of October 2026
Professor at UC San Diego
He is a professor at the Halıcıoğlu Data Science Institute at UC San Diego and an Amazon Scholar.
-
As of October 2026
Editing a SIAM journal
He is editor-in-chief of the SIAM Journal on Mathematics of Data Science.
Sources and credits
Photo: We have no published photo of Mikhail Belkin yet. Send us a free-licensed photo.
- Laplacian Eigenmaps and Spectral Techniques for Embedding and Clustering
- Broad Institute - Eric and Wendy Schmidt Center
- Mikhail Belkin - The Mathematics Genealogy Project
- Professor Mikhail Belkin published in PNAS | Computer Science and Engineering
- [1812.11118] Reconciling modern machine learning practice and the bias-variance trade-off
- UC San Diego Computer and Data Scientists Named to ACM’s Latest Class of Fellows