Every week, the OptML group gathers to update each other about their recent work and discuss various problems. In these one hour meetings, one student or faculty member gives a presentation, either to describe what they have learned in the past week or discuss the recent challenges they have been facing. The presentation is interactive with all members of the group participating, asking questions, and offering ideas. The group also occasionally invites researchers from industry or academia to present.

2021 Spring

04/20/2021 Tao Li OptML Student
Canonical Capsules: Unsupervised Capsules in Canonical Pose  paper1 paper2
04/14/2021 Yutong Dai OptML Student
Adversarial Robustness and Model Compression  paper1 paper2
04/07/2021 Liyuan Cao OptML Student
Zeroth-Order Regularized Optimization (ZORO): Approximately Sparse Gradients and Adaptive Sampling paper
03/31/2021 Suyun Liu OptML Student
Stochastic Alternating Balance Fair k-means and Alternating Bi-Objective Gradient Descent
03/24/2021 Shihong Xie OptML Guest
Interpretable, Robust, and Fair Learning on Graphs
03/10/2021 Tommaso Giovannelli, Oumaima Sohab OptML Student
Ridge Functions: Exploiting Lower-dimensional Structure paper1 paper2
03/03/2021 Qi Wang OptML Student
Deep Neural Networks as Gaussian Processes paper1 paper2
02/24/2021 Minhan Li OptML Student
On Regularization and Active-set Methods with Complexity for Constrained Optimization paper
02/17/2021 Tommaso Giovannelli Visiting Student
Derivative-free Methods for Mixed-integer Nonsmooth Constrained Optimization Problems paper1 paper2
02/10/2021 Michael O’Neill OptML PostDoc
Analysis of the BFGS Method with Errors paper1 paper2
02/03/2021 Baoyu Zhou OptML Student
Constrained and Composite Optimization via Adaptive Sampling Methods paper1 paper2

2020 Fall

12/09/2020 Minhan Li OptML Student
A High Probability Analysis of Adaptive SGD with Momentum paper
12/02/2020 Zheng Shi OptML Student
Neural Network Pruning paper1 paper2 paper3 paper4
11/18/2020 Tao Li OptML Student
Adaptive Attention via A Visual Sentinel for Image Captioning paper1 paper2
11/04/2020 Liyuan Cao OptML Student
Cubic Regularization of Newton Method and its Global Performance paper
10/28/2020 Sergey Rusakov OptML Student
Transformers: Architecture and Results paper1 paper2
10/21/2020 Oumaima Sohab OptML Student
Estimate Noise in Derivative-Free Optimization paper1 paper2 paper3
10/14/2020 Suyun Liu OptML Student
Fair Clustering Algorithms paper1 paper2 paper3
10/07/2020 Yuan Zeng OptML Guest
Legendre Memory Units: Continuous-Time Representation in Recurrent Neural Network paper
09/30/2020 Yutong Dai OptML Student
Active Set Identification paper
09/23/2020 Mertcan Yetkin OptML Student
Meta-learning: Basics and Recent Advancements slides
09/16/2020 Michael O’Neill OptML PostDoc
Worst Case Complexity paper1 paper2
09/09/2020 Tommaso Giovannelli Visiting Student
A Black-Box Optimization Approach for Emergency Department paper1 paper2 paper3 paper4
09/02/2020 Baoyu Zhou OptML Student
Nonsmooth BFGS Method paper1 paper2 paper3 paper4

2020 Spring

04/29/2020 Oumaima Sohab OptML Student
Direct Search Based on Probabilistic Descent paper
04/22/2020 Zheng Shi OptML Student
Residual Learning, Attention Mechanism and Multi-tasks Learning Networks slides
04/15/2020 Mohammad Pirhooshyaran OptML Student
Quantum Neuron paper
04/08/2020 Tao Li OptML Student
Can Deep Reinforcement Learning Improve Inventory Management? paper
04/01/2020 Suyun Liu OptML Student
A Review of Multi-Objective Optimization: Theory and Algorithms slides
03/25/2020 Minhan Li OptML Student
Stochastic Model-based Minimization of Weakly Convex functions paper
03/18/2020 Xin Shi OptML Student
A Survey of Recent Scalability Improvements for Semidefinite Programming paper
03/04/2020 Mertcan Yetkin OptML Student
Neural Architecture Search (NAS) Series 3 paper1 paper2
02/26/2020 Rusakov Sergey OptML Student
Neural Architecture Search (NAS) Series 2 paper1 paper2
02/19/2020 Majid Jahani OptML Student
Differentiable Neural ARchiTecture Search (DARTS) paper
02/12/2020 Baoyu Zhou OptML Student
Manifold Sampling for L1 Nonconvex Optimization paper
02/05/2020 Liyuan Cao OptML Student
Lagrange Quadratic Interpolation paper slides
01/29/2020 Haidong Gu OptML Student
A Decent Algorithm and a Homotopy Method for Solving Lasso Problem paper
01/22/2020 Rui Shi OptML Student
Unified Convergence Analysis of Stochastic Momentum Methods for Convex and Non-convex Optimization paper

2019 Fall

11/13/2019 Xin Shi, Minhan Li OptML Student
Adversarial Machine Learning Algorithms slides
11/06/2019 Xin Shi, Minhan Li OptML Student
Introduction to Adversarial Machine Learning slides
10/30/2019 Suyun Liu OptML Student
Online Convex Optimization in the Bandit Setting slides
Frank E. Curtis OptML Faculty
Systematic Insights on the Fisher Matrix and Comments on paper
10/16/2019 Rui Shi OptML Student
Recent papers regarding Online Learning paper1 paper2
Majid Jahani, Baoyu Zhou OptML Student
Second-Order Methods for Deep Learning paper
10/09/2019 Tao Li OptML Student
First Order Methods for Online Convex Optimization  slides
Majid Jahani, Baoyu Zhou OptML Student
Second-Order Methods for Deep Learning paper
10/02/2019 Lili Song OptML Student
Introduction to Online Convex Optimization  slides
Majid Jahani, Baoyu Zhou OptML Student
Second-Order Methods for Deep Learning slides
09/25/2019 Rusakov Sergey OptML Student
Approaches to Solving Semantic Segmentation slides
09/18/2019 Mertcan Yetkin, Mohammad Pirhooshyaran OptML Student
Convolutional Neural Network (CNN):
Basics and Recent Advancements slides
09/11/2019 Liyuan Cao, Haidong Gu OptML Student
Introduction to Object Detection slides
09/04/2019 Liyuan Cao, Haidong Gu OptML Student
Computer Vision Tutorial slides


2019 Spring

04/17/2019 Mertcan Yetkin OptML Student
Generative Adversarial Nets paper
04/10/2019 Suyun Liu OptML Student
Stochastic Gradient Methods for Non-Smooth
Non-Convex Regularized Optimization paper
04/03/2019 Albert Berahas OptML PostDoc
Analysis of the BFGS method with errors paper
03/20/2019 Minhan Li OptML Student
Decentralized Quasi-Newton Methods paper
03/06/2019 Liyuan Cao OptML Student
Random Gradient-Free Minimization of Convex
Functions paper
02/20/2019 Baoyu Zhou OptML Student
A Newton-Based Method for Nonconvex Optimization with Fast Evasion of Saddle Points paper
02/13/2019 Rui Shi OptML Student
Theoretical results in Online Learning paper
02/06/2019 Majid Jahani OptML Student
Catalyst Acceleration for First-order Convex Optimization:
from Theory to Practice paper


12/05/2018 Mohammad Pirhooshyaran OptML Student
Adaptive Cubic Regularization paper
11/14/2018 Liyuan Cao OptML Student
Natural Evolutionary Strategies paper
11/07/2018 Minhan Li OptML Student
Stochastic Quasi Newton Methods paper
10/31/2018 Frank E. Curtis OptML Faculty
Theory of BFGS paper
10/24/2018 Suyun Liu OptML Student
Complexity of gradient descent for multiobjective optimization paper
10/17/2018 Sarper Aydin OptML Student
A derivative-free trust-region algorithm for the optimization of functions smoothed via Gaussian convolution using adaptive multiple importance sampling paper
10/10/2018 Albert Berahas OptML Postdoc
Adaptive Sampling Strategies for Stochastic Optimization paper
09/26/2018 Baoyu Zhou OptML Student
How to Escape Saddle Points Efficiently paper
09/19/2018 Mertcan Yetkin OptML Student
ADMM and Accelerated ADMM as Continuous Dynamical Systems paper
09/05/2018 Rui Shi OptML Student
Introduction to Gradient Boosting Models book
05/02/2018 Quoc Tran-Dinh visitor
Generalized Self-Concordant Functions: A Recipe for Newton-Type Methods paper
04/18/2018 Mohammad Pirhooshyaran OptML Student
Stochastic cubic regularization for fast nonconvex optimization paper
04/11/2018 Nicolas Loizou Visiting Student from University of Edinburgh
Momentum and Stochastic Momentum for Stochastic Gradient, Newton, Proximal Point and Subspace Descent Methods paper
04/04/2018 Rui Shi OptML Student
Stochastic Trust Region Algorithm
03/28/2018 Lam Nguyen OptML Student
Stability of stochastic gradient descent paper
03/08/2018 Majid Jahani OptML Student
Efficient Distributed Hessian Free Algorithm for Large-scale Empirical Risk Minimization via Accumulating Sample Strategy
02/28/2018 Majid Jahani OptML Student
Adaptive Newton Method for Empirical Risk Minimization to Statistical Accuracy paper
02/21/2018 Katya Scheinberg OptML Faculty
Global convergence rate analysis of unconstrained optimization
methods based on probabilistic models paper
02/14/2018 Courtney Paquette OptML PostDoc Researcher
Local search for nonsmooth and nonconvex problems paper
02/07/2018 Mohammadreza Samadi OptML student
Recent Updates on Subsampled Newton methods
01/31/2018 All members  
Group discussion in the new office


12/03/2017 Mohammadreza Nazari OptML student
Policy gradients in application to reinforcement learning
11/17/2017 Albert Berahas visitor
A Multi-Batch L-BFGS Method for Machine Learning
11/10/2017 Lam Nguyen OptML student
When do stochastic gradient algorithms work well for training deep neural networks
11/03/2017 Francesco Orabona visitor
Coin Betting for Backprop without Learning Rates and More
10/20/2017 Chenxin Ma OptML student
Introduction to natural gradient descent II
10/13/2017 Rui Shi OptML student
Different measures on convergence results
10/06/2017 Chenxin Ma OptML student
Introduction to natural gradient descent I
09/22/2017 Courtney Paquette visitor
Acceleration for Gradient-Based Non-Convex Optimization
09/15/2017 Xi He OptML student
TensorFlow tutorial II
09/08/2017 Xi He OptML student
TensorFlow tutorial I
05/02/2017 Matt Menickelly OptML student
On graduated optimization for stochastic non-convex problems
04/18/2017 Majid Jahani OptML student
Finite-sum Composition Optimization via Variance Reduced
Gradient Descent
04/11/2017 Mertcan Yetkin OptML student
SVM for solving optimal control
04/04/2017 Liyuan Cao OptML student
k-Components: A Method for Analyzing Data in Radial Shape
03/28/2017 Mohammadreza Samadi OptML student
Gradient Descent with Momentum
03/21/2017 Jie Liu OptML student
SARAH: part II
03/07/2017 Lam Nguyen OptML student
SARAH: A Novel Method for Machine Learning Problems Using Stochastic Recursive Gradient
02/28/2017 Hiva Ghanbari OptML student
A Globally Convergent Inexact Newton Method for Systems of Monotone Equations
02/21/2017 Chenxin Ma OptML student
Underestimated Sequence Part II
02/14/2017 Xi He OptML student
Sub-Sampled Exact and Inexact Newton Methods
02/07/2017 Wei Guo OptML student
Barzilai-Borwein Step Size for Stochastic Gradient Descent paper
02/01/2017 Majid Jahani OptML student
Underestimated Sequence
11/29/2016 Matt Menickelly OptML student
Sparsity-Constrained Gaussian Graphical Models for Anomaly Detection
11/22/2016 Jie Liu OptML student
Fast stochastic methods for nonsmooth nonconvex optimization paper
11/01/2016 Mohammadreza Samadi OptML student
Proof technique in non-convex optimization
10/25/2016 Rui Shi OptML student
Robust Stochastic Approximation Approach to Stochastic Programming
10/18/2016 N/A OptML student
Discussions with Mark Schmidt
10/04/2016 Kürşat Kemikli OptML student
A limited memory quasi-Newton trust-region method for box constrained optimizatio
09/27/2016 Hiva Ghanbari OptML student
A Universal Catalyst for First-Order Optimization paper
09/20/2016 Chenxin Ma OptML student
An optimal first order method based on optimal quadratic averaging paper
09/13/2016 Wei Guo OptML student
Handling Nonpositive Curvature in a
Limited Memory Steepest Descent Method
08/30/2016 Xi He OptML student
Second order methods in training deep neural network
05/05/2016 Martin Takáč OptML faculty
Theano And GPU computing in Python: Code
03/31/2016 Mohammadreza Samadi OptML student
Efficient Trust Region Subproblem Solvers
03/24/2016 Chenxin Ma OptML student
Second Order Stochastic Optimization in Linear Time
03/10/2016 Matt Menickelly OptML student
Random Sampling in Stochastic Optimization
03/03/2016 Xi He OptML student
Intro to Deep Neural Networks
02/25/2016 Milad Siami OptML guest
System Performance Measures for Noisy Consensus Networks
02/18/2016 Jiawei Zhang OptML guest
Deep Neural Network Learning Artistic Style
02/11/2016 Milad Siami OptML guest
System Performance Measures for Noisy Consensus Networks
02/04/2016 Jie Liu OptML student
Hybrid Optimization Methods in AC Optimal Power Flow
01/28/2016 Frank E. Curtis OptML faculty
Theory of Stochastic Gradient Methods


12/02/2015 Hiva Ghanbari OptML student
12/02/2015 Mohammadreza Samadi OptML student
11/25/2015 Xi He OptML student
11/11/2015 Francesco Orabona Yahoo! Labs
10/28/2015 Chenxin Ma OptML student
10/21/2015 Jeffrey Larson Argonne National Laboratory
10/21/2015 Matt Menickelly OptML student
10/14/2015 Wei Guo OptML student
10/07/2015 Jie Liu OptML student
09/30/2015 Hiva Ghanbari OptML student
09/30/2015 Mohammadreza Samadi OptML student
09/16/2015 Rui Shi OptML student
09/16/2015 Matt Menickelly OptML student
09/09/2015 Xi He OptML student
09/09/2015 Chenxin Ma OptML student
09/02/2015 Wei Guo OptML student
09/02/2015 Jie Liu OptML student
08/26/2015 Hiva Ghanbari OptML student
08/26/2015 Mohammadreza Samadi OptML student
08/19/2015 Xi He OptML student

Reading Seminars (2017 and earlier)


Date and Location Speaker Paper or Topic
2017/09/06, 2:30pm, Mohler 121 Xi He
  • Escape of saddle points for non-convex optimization
2017/04/11, 2:00pm, Mohler 121 Xi He
2017/04/04, 2:00pm, Mohler 121 Chenxin Ma
2017/02/01, 2:30pm, Mohler 121 Lam Nguyen
  • His new research with Jie Liu
2017/01/17, 3:15pm, Mohler 375 Lam Nguyen
2017/01/10, 3:15pm, Mohler 375 Xi He

Fall 2016

Date and Location Speaker Paper or Topic
2016/12/20, 2:30pm, Mohler 375 Chenxin Ma
2016/10/04, 2:30pm, Mohler 375 Xi He
2016/10/04, 2:30pm, Mohler 375 Chenxin Ma
2016/09/27, 2:30pm, Mohler 375 Jie Liu
2016/09/27, 2:30pm, Mohler 375 Xi He
2016/08/31, 9:00am, Mohler 375 Chenxin Ma

Fall 2015

Date and Location Speaker Paper or Topic
2015/08/28, 4:11pm, Mohler 451
Xi He
2015/09/04, 4:11pm, Mohler 451 Jie Liu
2015/09/11, 4:11pm, Mohler 451 Xi He 

Chenxin Ma

2015/09/18, 10:00pm, Mohler 375
Jie Liu
  • CRF – tutorial
2015/09/25, 4:11pm, Mohler 451 Xi He 

Chenxin Ma

2015/12/22, 12:00pm, Mohler 375 Chenxin Ma 

Jie Liu

  • Sparse Linear Programming via Primal and Dual Augmented Coordinate Descent
  • Solving Random Quadratic Systems of Equations Is Nearly as Easy as Solving Linear Systems
2015/12/27, 12:00pm, Mohler 375 Jie Liu 

Chenxin Ma

  • Scaling Up Distributed Stochastic Gradient Descent Using Variance Reduction