The University of Texas at Austin
Department of Electrical and Computer Engineering
EE381V-11: Large Scale Optimization Fall 2012
Problem Set Two Solutions
Caramanis/Sanghavi
Due: Thursday, September 20, 2012.
Matlab and Computational Assignments Please p
The University of Texas at Austin
Department of Electrical and Computer Engineering
EE381V-11: Large Scale Optimization Fall 2012
Problem Set Three Solutions
Caramanis/Sanghavi
Due: Thursday, September 27, 2012.
Written Problems
1. Gradient descent with d
The University of Texas at Austin
Department of Electrical and Computer Engineering
EE381V-11: Large Scale Optimization Fall 2012
Problem Set One Solutions
Caramanis/Sanghavi
Due: Thursday, September 13, 2012.
Matlab and Computational Assignments.
1. This
The University of Texas at Austin
Department of Electrical and Computer Engineering
EE381V-11: Large Scale Optimization Fall 2012
Problem Set Seven Solutions
Caramanis/Sanghavi
Due: Thursday, November 8, 2012.
Written Problems
1. Show that sub gradients h
The University of Texas at Austin
Department of Electrical and Computer Engineering
EE381V-11: Large Scale Optimization Fall 2012
Problem Set Nine Solutions
Caramanis/Sanghavi
Due: Thursday, November 29, 2012.
Matlab and Computational Assignments. Please
EE381V: Machine Learning for Large-Scale Data
Lecturer: Alexandros G. Dimakis
March 22, 2016
Scribe: Amir Gholami, Chad Voegele
Yunan Yang, Cagdas Yelen
Lecture 17: Submodularity and Feature Selection
1
Introduction to Feature Selection
In this lecture we
EE381V: Machine Learning: Large-Scale Data
Feb 11th, 2016
Lecture: Convex Optimization
Lecturer: Constantine Caramanis
1
1.1
Scribe: Tianyang Li, Rongting Zhang
Yanyao Shen, Arun Sai Suggala
Last Time
Gradient Descent
Gradient descent is a popular method
EE381V- Machine Learning for Large-Scale Data
Jan 28th, 2016
Lecture 4: LSH and JL Embedding
Lecturer: Alexandros G. Dimakis,
Constantine Caramanis
1
Scribe: Kiyeon Jeon, Muryong Kim,
Taewan Kim, Tzu-Ling Kan
Locality sensitive hashing
In this lecture, we
EE381V- Machine Learning for Large-Scale Data
Feb 4th, 2016
Lecture 6: Subspace Embeddings and Fast JL
Lecturer: Constantine Caramanis
1
Scribe: Ger Yang, Heng-Lu Chang,
Che-Chun Lin, Zhongqi Wang
Review
Last week we proofed Johnson-Lindenstrauss (JL) the
EE381V: Machine Learning for Large-Scale Data
March 1st, 2016
Lecture 13: Multiplicative Weights Update Methods
Lecturer: Alexandros G. Dimakis
Constantine Caramanis
1
Scribe: Qi Lei, Biying Xu,
Libo Chen, Kevin Wang
About this lecture
In last several lec
EE381V: Machine Learning for Large Scale Data
Mar 24th, 2016
Lecture 18: Sparse Linear Regression and AUC
Lecturer: Alexandros G. Dimakis
1
Scribe: Xiaoxia Wu, Yuyang Xie, Yuming Sheng
Introduction
In this lecture, we will study sparse regression. Specifi
EE381V: Machine Learning for Large-Scale Data
February 23, 2016
Lecture 11: Boosting
Lecturer: Alexandros G. Dimakis
Scribe: Ashish Bora, Surbhi Goel, Jessica Hoffmann, John Kallaugher
1
Introduction
This lecture will deal with the question of how to cons
EE381V: Machine Learning: Large Scale Data
Lecturer: Alexandros G. Dimakis
1
Feb 16th, 2016
Scribes: Ethan R. Elenberg, Erik Lindgren
Statistical Learning Framework
The following general framework for statistical learning closely follows [?]:
Domain Set
EE381V: Large-Scale Machine Learning
8 March 2016
Lecture 15: Support Vector Machines, The Kernel Trick, and PEGASUS
Lecturer: Constantine Caramanis Scribe: Sid Kapur, Yu-Sian Jiang,Daqi Xu, Zhiyuan Zou
1
Synopsis
This lecture will be about classification
EE381V: Machine Learning
Feb 6th, 2016
Lecture 7: Convex Optimization Review
Lecturer: Constantine Caramanis
Scribes: Rohith Prakash, Michael Bartling, Henry Chen, Rahi Kalantari
1
Previous Lecture
In the last lecture, we continued our discussion on Johns
EE381V: Machine Lrn: Lrg-Scale Data
Mar 31th, 2016
Lecture 20:
Lecturer: Alex Dimakis & Constantine Caramanis
Scribe: Yitao Chen, Xiaodan Xi, Mengshi Zhang, Qi Wang
1
Alexs ideas about the Kaggle homework
There are some key ideas for getting good performa
EE381V: Large Scale Machine Learning
Mar 10th, 2016
Lecture 16: Submodularity
Lecturer: Alexandros G. Dimakis
1
Scribe: Shashank, Phanindra, Ankita, Soumya
Introduction
The concept of submodularity of set functions has widespread applications in the field
EE381V: Large ML
Feb 18th, 2016
Lecture 10: Linear predictor: classifier and regression
Lecturer: Alexandros G. Dimakis
1
Scribe: J. Zhang, T. Zhou, H. Dong, M. Jin
Introduction
In the last lecture we discussed statistical learning framework, empirical ri
The University of Texas at Austin
Department of Electrical and Computer Engineering
EE381V-11: Large Scale Optimization Fall 2012
Problem Set Eight Solutions
Caramanis/Sanghavi
Due: Thursday, November 16, 2012.
Written Problems
1. Consider the
1 -regulari
The University of Texas at Austin
Department of Electrical and Computer Engineering
EE381V-11: Large Scale Optimization Fall 2012
Problem Set Six Solutions
Caramanis/Sanghavi
Due: Thursday, November 1, 2012.
Reading Assignments
1. Reading: Boyd & Vandenbe
The University of Texas at Austin
Department of Electrical and Computer Engineering
EE381V-11: Large Scale Optimization Fall 2012
Problem Set Five Solutions
Caramanis/Sanghavi
Due: Thursday, October 18, 2012.
Written Problems
1. For A an m n matrix, and b
The University of Texas at Austin
Department of Electrical and Computer Engineering
EE381V-11: Large Scale Optimization Fall 2012
Problem Set Four Solutions
Caramanis/Sanghavi
Due: Thursday, October 4, 2012.
Matlab and Computational Assignments. Please pr
The University of Texas at Austin
Department of Electrical and Computer Engineering
EE380K: Linear Systems TheoryFall 2010
Problem Set Zero Solutions
Constantine Caramanis
Due: Wednesday, September 1, 2010.
Matlab and Computational Assignments
3.
Algorit
EE 381V: Large Scale Optimization
Fall 2012
Lecture 26 December 4
Lecturer: Caramanis & Sanghavi
26.1
Scribe: Tao Huang & Arda Sisbot
Introduction
This lecture will cover a specic algorithm for parallelizing sub-gradient descent. The result
is that any su
EE 381V: Large Scale Optimization
Fall 2012
Lecture 5 September 13
Lecturer: Caramanis & Sanghavi
5.1
Scribe: Debarati Kundu & Tejaswini Ganapathi
Topics covered
Recap of denitions and theorems taught in the previous lecture
Coordinate Descent Method
S
EE 381V: Large Scale Optimization
Fall 2012
Lecture 4 September 11
Lecturer: Caramanis & Sanghavi
4.1
Scribe: Gezheng Wen, Li Fan
Gradient Descent
The idea relies on the fact that f (x(k) ) is a descent direction.
Algorithm description
x(k+1) = x(k) (k) f
EE 381V: Large Scale Optimization
Fall 2012
Lecture 3 September 06
Lecturer: Caramanis & Sanghavi
3.1
Scribe: Samer Chucri & Somsubhra Barik
Topics covered
Projection onto a Convex Set
Separation of Convex Sets
Unconstrained Optimization : Gradient Des
Dealing with Massive Data
January 31, 2011
Lecture 2: Distinct Element Counting
Lecturer: Sergei Vassilvitskii
1
Scribe:Ido Rosen & Yoonji Shin
Introduction
We begin by defining the stream formally.
Definition 1 A finite data stream, X = x1 x2 x3 .xn , is
EE381V: Machine Learning: Large Scale Data
Jan 26th, 2016
Lecture 3: Streaming and Sketching
1
Lecturer: Alexandros G. Dimakis
Scribe: Scott Johnston, Xueyu Mao
Constantine Caramanis
Pranav Madadi, Vatsal Shah
Introduction
We have learned universal hashin
EE381V: Machine Learning for Large-Scale Data
Jan 19th, 2016
Lecture 1: Hashing and Streaming
Lecturer: Alexandros G. Dimakis
1
Scribe:
Abolfazl Hashemi, Anum Ali
Ali Khodabakhsh, Isfar Tariq
Introduction
In this course we are interested in algorithms for