Mehryar Mohri

About

The provided page text does not contain a detailed professional biography or research focus of Professor Mehryar Mohri. It primarily lists current and former members of his research group, along with their titles and affiliations. Therefore, there is no specific biographical information or description of his research contributions available in the given content.

Research topics

• Computer Science
• Artificial Intelligence
• Machine Learning
• Data Mining
• Data science
• World Wide Web
• Engineering
• Mathematics
• Mathematical optimization
• Distributed computing

Selected publications

• A Field Guide to Federated Optimization

  arXiv (Cornell University) · 2021 · 167 citations

  Federated learning and analytics are a distributed approach for collaboratively learning models (or statistics) from decentralized data, motivated by and designed for privacy protection. The distributed learning process can be formulated as solving federated optimization problems, which emphasize communication efficiency, data heterogeneity, compatibility with privacy and system requirements, and other constraints that are not primary considerations in other problem settings. This paper provides…

  DOI

• Guarantees for Epsilon-Greedy Reinforcement Learning with Function Approximation

  arXiv (Cornell University) · 2022 · 26 citations

  Myopic exploration policies such as epsilon-greedy, softmax, or Gaussian noise fail to explore efficiently in some reinforcement learning tasks and yet, they perform well in many others. In fact, in practice, they are often selected as the top choices, due to their simplicity. But, for what tasks do such policies succeed? Can we give theoretical guarantees for their favorable performance? These crucial questions have been scarcely investigated, despite the prominent practical importance of these…

  PublisherOA PDFDOI

• Generalized Distributional Alignment Games for Unbiased Answer-Level Fine-Tuning

  arXiv (Cornell University) · 2026-05-04

  preprintOpen access1st authorCorresponding

  The Distributional Alignment Game framework provides a powerful variational perspective on Answer-Level Fine-Tuning (ALFT). However, standard algorithms for these games rely on estimating logarithmic rewards from small batches, introducing a systematic bias due to Jensen's inequality that can destabilize training. In this paper, we systematically resolve this structural estimation bias. First, we generalize the alignment game to arbitrary Bregman divergences, showing that for a family of geometr…

  PublisherDOI

• Linear-Core Surrogates: Smooth Loss Functions with Linear Rates for Classification and Structured Prediction

  ArXiv.org · 2026-04-30

  articleOpen access1st authorCorresponding

  The choice of loss function in classification involves a fundamental trade-off: smooth losses (like Cross-Entropy) enable fast optimization rates but yield slow square-root consistency bounds, while piecewise-linear losses (like Hinge) offer fast linear consistency rates but suffer from non-differentiability. We propose Linear-Core (LC) Surrogates, a new family of convex loss functions that resolve this tension by stitching a linear core to a smooth tail. We prove that these surrogates are diffe…

  PublisherOA PDF

• Balancing fairness: learning with conflicting objectives and user feedback

  Annals of Mathematics and Artificial Intelligence · 2026-03-12

  articleOpen accessSenior author

  Large-scale learning systems are often evaluated across multiple, sometimes conflicting, objectives. How can we effectively learn and optimize such complex systems with potentially incompatible goals? Moreover, how do we improve these systems when user feedback, potentially highlighting previously unaddressed issues, becomes available? We propose a novel theoretical model for learning and optimizing such systems. Instead of relying on a static or predefined trade-off among objectives, our model…

  PublisherOA PDFDOI

Recent grants

• RI: Small: Ensemble Methods for Structured Prediction

  NSF · $407k · 2011–2016

• AitF: FULL: Collaborative Research: PEARL: Perceptual Adaptive Representation Learning in the Wild

  NSF · $400k · 2015–2021

• RI: Small: Collaborative Research: On-Line Learning Algorithms for Path Experts with Non-Additive Losses

  NSF · $275k · 2016–2021

Frequent coauthors

• Corinna Cortes

  Google (United States)

  148 shared

• Afshin Rostamizadeh

  54 shared

• Michael Riley

  40 shared

• Cyril Allauzen

  Google (United States)

  37 shared

• Vitaly Kuznetsov

  Odessa National Economics University

  35 shared

• Ananda Theertha Suresh

  35 shared

• Scott Cheng‐Hsin Yang

  26 shared

• Yishay Mansour

  23 shared

Labs

• Mehryar Mohri's Research GroupPI

  Research Group@Courant

Awards & honors

• Best Paper Award EC 2025
• NeurIPS 2025 Oral

Similar researchers at New York University

• Rob Fergush-95
• Dennis Shashah-84
• Nasir Memonh-70