About me

I am a postdoctoral associate at MIT Sloan. I received my Ph.D. in the multidisciplinary program Algorithms, Combinatorics and Optimization / Computer Science (ACO/CS) at Georgia Tech. I was very fortunate to be advised by Santosh Vempala. I have completed my M.Sc. in Computer Science at the University of British Columbia, under supervision of Bruce Shepherd and Mark Schmidt. Previously, I received my B.Sc. in Computer Engineering Department from Sharif University of Technology.
My research is in optimization and areas related to optimization, including machine learning and algorithms.

Research Interests

• Combinatorial and Convex Optimization
• Design and Analysis of Algorithms, Approximation and Randomized Algorithms
• Theoretical Machine Learning, Theoretical Aspects of Deep Learning, Tensor Decomposition
• Machine Learning, Fairness, Diffrerential Privacy, Data Mining

Work Experience

• Postdoctoral Associate MIT Sloan. September 2023 - Present
  Host: Swati Gupta.
• Research Scientist Intern in Real-time Experiences and Algorithms Lab at Adobe. May 2022 - August 2022
  Hosts: Anup Rao, Tung Mai, and David Arbour.
  Worked on experimental design, average treatment effect (ATE) and individual treatment effect (ITE).
• Student researcher in Algorithms and Optimization Team at Google. August 2021 - May 2022
  Hosts: Matthew Fahrbach and Thomas Fu.
  Worked on tensor decomposition methods for deep neural network and feature engineeting applications.
• Research intern in Algorithms and Optimization Team at Google. May 2021 - August 2021
  Hosts: Matthew Fahrbach and Thomas Fu.
  Worked on tensor decomposition methods for deep neural network applications.

Education

• Ph.D. in ACO/CS 2019 - 2023
  Algorithms, Combinatorics, and Optimization / Computer Science, Georgia Institute of Technology
  Advisor: Santosh Vempala
  Minor: Algebraic and Enumerative Combinatorics.
• M.Sc. in Computer Science 2017 - 2019
  Department of Computer Science, University of British Columbia
  Advisors: Bruce Shepherd and Mark Schmidt
  Thesis title: "Beyond Submodular Maximization: One-Sided Smoothness and Meta-Submodularity"
• B.Sc. in Computer Engineering 2011 - 2016
  Department of Computer Engineering, Sharif University of Technology
  Advisor: Hamid Zarrabi-Zadeh
  Thesis title: "Discrete Voronoi Game"

Awards and Honors

• The Machine Learning Center at Georgia Tech (ML@GT) Fellowship. 2021
• Institute for Data Engineering and Science (IDEaS) and Transdisciplinary Research Institute for Advancing Data Science (TRIAD) Research Scholarship for Ph.D. Students and Postdocs. 2020
• Borealis AI Global Fellowship Award 2018
• Silver Medal at Iranian National Mathematical Olympiad 2010

Publications

• The Bit Complexity of Efficient Continuous Optimization FOCS 2023
  M. Ghadiri, R. Peng and S. Vempala
  pdf
• On Symmetric Factorizations of Hankel Matrices FOCS 2023
  M. Ghadiri
  pdf
• Approximately Optimal Core Shapes for Tensor Decompositions ICML 2023
  M. Ghadiri, M. Fahrbach, G. Fu, and V. Mirrokni
  pdf
• Subquadratic Kronecker Regression with Applications to Tensor Decomposition NeurIPS 2022
  M. Fahrbach, G. Fu, and M. Ghadiri (Listed Alphabetically)
  pdf
• Amortized rejection sampling in universal probabilistic programming AISTATS 2022
  S. Naderiparizi, A. Scibior, A. Munk, M. Ghadiri, A. G. Baydin, B. G. Hansen, C. S. de Witt, R. Zinkov, P. Torr, T. Rainforth, Y. W. Teh, F. Wood
  [Oral Presentation]
  A preliminary version appeared in PROBPROG 2020.
  pdf
• Socially Fair k-Means Clustering FAccT 2021
  M. Ghadiri, S. Samadi, and S. Vempala
  pdf
• Beyond Submodular Maximization via One-Sided Smoothness SODA 2021
  M. Ghadiri, R. Santiago, and B. Shepherd
  pdf
• Distributed Maximization of Submodular Plus Diversity Functions for Multi-label Feature Selection on Huge Datasets AISTATS 2019
  M. Ghadiri and M. Schmidt
  pdf
• Scalable Feature Selection via Distributed Diversity Maximization AAAI 2017
  M. Ghadiri^*, S. Abbasi Zadeh^*, M. Zadimoghaddam and V. Mirrokni (* equal contribution)
  [Oral Presentation]
  pdf
• A Multiscale Agent-Based Framework Integrated with a Constraint-Based Metabolic Network Model of Cancer
  for Simulating Tumor Growth Molecular BioSystems Journal 2017
  M. Ghadiri^*, M. Heidari^*, S. H. Mousavi and S. A. Marashi (* equal contribution)
  
  Link
• Linear Relaxations for Finding Diverse Elements in Metric Spaces NIPS 2016
  A. Bhaskara, M. Ghadiri, V. Mirrokni and O. Svensson
  
  pdf
• Minimizing the Total Movement for Movement to Independence Problem on a Line CCCG 2016
  M. Ghadiri and S. Yazdanbod
  
  pdf
• Active Distance-Based Clustering using K-medoids PAKDD 2016
  M. Ghadiri^*, A. Aghaee^*, M. Soleymani Baghshah (* equal contribution)
  
  pdf

Working Papers

• Constant-Factor Approximation for Socially Fair k-Clustering 2021
  M. Ghadiri, M. Singh and S. Vempala
  pdf
• A Parameterized Family of Meta-Submodular Functions 2020
  M. Ghadiri, R. Santiago, and B. Shepherd
  pdf

Workshops

• Efficient Bayesian Inference for Nested Simulators AABI 2019
  B. G. Hansen, C. S. de Witt, R. Zinkov, S. Naderiparizi, A. Scibior, A. Munk, F. Wood, M. Ghadiri, P. Torr, Y. W. Teh, A. G. Baydin, T. Rainforth
  
  pdf