Statistical Learning and Data Mining

Fall 2014

All meetings will be held every other Tuesday between 11:10 am - 12:30 pm in CH212.
We will start with Optimization for Statistics as one of the themes of Fall 2014.
Presentation and discussion schedule will be posted as we go along.

• September 9, 2014
  Speaker: Andrew Landgraf
  Title: Data Science for Social Good
  


• September 23, 2014
  Speakers: Vince Vu, Yoon Lee, and Yunzhang Zhu
  Title: Selected Examples of Optimization from Lange et al. (2014)
  
  Reading groups
  

  Newton:	Qian Qian, Abhijoy Saha, Shanshan Tu, Weiyi Xie, Jiaqi (Jessie) Zaetz*, Sebastian Kurtek
  MM:	Andrew Landgraf*, Yi Lu, Huong (Sophie) Nguyen, Ran Wei, Yuan Zhang, Yoon Lee
  Lagrange:	Liubo Li*, Yinghao Sun, Zhifei Yan, Vince Vu, Yunzhang Zhu

• October 7, 2014
  
  Weiyi Xie & Jessie Zaetz: Quasi-Newton Methods
  Yi Lu: Tricks of the Trade from A Tutorial on MM Algorithms
  Liubo Li: Introduction to ADMM algorithm
  
  
• October 21, 2014
  Title: Comparing Optimization Techniques in the NBA Team Strength Example from A Tutorial on MM Algorithms
  Discussants: Sophie Nguyen, Qian Qian, Shanshan Tu, and Zhifei Yan
  Data


• November 4, 2014
  No Meeting
  


• November 18, 2014
  Jessie Zaetz: Conjugate Gradient Method
  Andrew Landgraf: Convergence of the MM Algorithm from Section 12.4 of Optimization
  Liubo Li: ADMM for the Lasso Problem from Chapter 6 of Distributed Optimization and Statistical Learning via the Alternating Direction Method of Multipliers
  


• December 2, 2014
  Abhijoy Saha: Non Linear Conjugate Gradient Method
  Yi Lu & Sophie Nguyen: MM Algorithm for Image Debluring from Adaptative Total Variation Image Deblurring: A Majorization-Minimization Approach
  Zhifei Yan: Dynamically Updating the Penalty Parameter (ρ) in ADMM