DS-GA 1013 / MATH-GA 2821 Optimization-based Data Analysis

Instructor: Carlos Fernandez-Granda (cfgranda@cims.nyu.edu)
Teaching Assistant: Brett Bernstein (brettb@cims.nyu.edu)

This course provides a unifying description of optimization-based methods designed to tackle data-analysis problems, including sparse regression, compressed sensing, super-resolution, matrix completion, clustering and manifold learning. We will analyze these techniques from a mathematical and algorithmic point of view and describe their application to a wide range of practical problems. See the schedule for more details.

Announcements

General Information

Prerequisites

Calculus, linear algebra and probability (at the level of the DS GA 1002 notes). Some programming skills and some exposure to statistics, machine learning or optimization are desirable.

Lecture

Wednesday 3:30-5:10 pm, 60 5th Ave (CDS) room 110

Recitation

Monday 5:20-6:10 pm, 60 5th Ave (CDS) room 110

Office hours

Carlos: Friday 5:30-6:30 pm, 60 5th Ave (CDS) room 606
Brett: Friday 4:00-5:00 pm, 60 5th Ave (CDS) room 620

Grading policy

Biweekly homework (50%) + Project proposal (10%) + Project (40%)

Homework

Homework deadlines are posted on the schedule. The assignments should be submitted as a pdf through NYU classes. Any code you write should be submitted in a zip file. The solutions and the grades will be available also on NYU classes.

Feel free to discuss the homework with other students in person or on Piazza, but do not share specific answers and make sure that you write your own personal solutions yourself. Always explain your thought process. If you use results from the notes or a book reference them adequately.

Books

We will provide self-contained notes. Some useful additional references are