PeopleofSuperintelligence

Researcher

Ali Rahimi

Ali Rahimi is a computer scientist whose work spans computer vision, kernel methods, and sensor-based activity recognition. He studied low-resolution vision at MIT and developed methods for unsupervised regression and time-series learning. With Benjamin Recht, he introduced random-feature methods for kernel machines and analyzed shallow randomized networks. Their 2007 paper received the 2017 NeurIPS Test of Time Award. In summer 2019, he served as a visiting scientist for the Simons Institute's Foundations of Deep Learning program.

Timeline

  1. 1997–2001

    Degrees at Berkeley and MIT

    At UC Berkeley, he earns a BS in EECS in 1997 and an MS in media arts and sciences at MIT in 2001. His Bug Vision project explores vision from sparsely sampled images.

  2. 2005

    Nonlinear video models

    In 2005, he co-authors an MIT technical report with Ben Recht and Trevor Darrell on nonlinear latent-variable models for video sequences.

  3. 2006

    Unsupervised regression

    With Ben Recht, he develops an unsupervised regression method in 2006 to estimate low-dimensional processes from high-dimensional observations without input-output examples.

  4. 2007

    Random features for kernels

    With Benjamin Recht, he proposes mapping data to random low-dimensional features so fast linear methods can scale kernel learning.

  5. 2008

    The Mobile Sensing Platform

    In 2008, he co-authors The Mobile Sensing Platform, a paper on embedded activity recognition.

  6. 2008

    Weighted sums of random kitchen sinks

    In 2008, he and Benjamin Recht analyze shallow random networks and derive test-error bounds based on dataset size and the number of randomized nonlinearities.

  7. 2017

    A Test of Time Award

    In 2017, the NeurIPS Test of Time Award recognizes his 2007 paper with Benjamin Recht, Random Features for Large-Scale Kernel Machines.

  8. 2019

    Visiting scientist at Simons

    In summer 2019, he serves as a visiting scientist for the Foundations of Deep Learning program at the Simons Institute.

Sources and credits

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

  1. Ali Rahimi vismod.media.mit.edu. Checked 15 October 2026.
  2. CSAIL Publications publications.csail.mit.edu. Checked 15 October 2026.
  3. Unsupervised Regression with Applications to Nonlinear System Identification proceedings.neurips.cc. Checked 15 October 2026.
  4. Random Features for Large-Scale Kernel Machines papers.neurips.cc. Checked 15 October 2026.
  5. The Mobile Sensing Platform: An Embedded Activity Recognition System research.google. Checked 15 October 2026.
  6. Weighted Sums of Random Kitchen Sinks: Replacing minimization with randomization in learning papers.nips.cc. Checked 15 October 2026.
  7. Awards neurips.cc. Checked 15 October 2026.
  8. Ali Rahimi simons.berkeley.edu. Checked 15 October 2026.