Fall 2019, Course Website: canvas.gatech.edu

Tuesday, Thursday 3:00 – 4:15 pm, Love (MRDC II) 185

Instructor: Prof. Swati Gupta, Office: Groseclose 437, Email: swatig@gatech.edu
Office hours: Wednesdays 4:00 to 5:00 pm in Groseclose 437 or by appointment via email. 

Teaching Assistants: 
Linxi Xiao (xiaolinxi@gatech.edu), 
Office hours: Monday 11:30AM-12:30 PM, Location: ISYE Main Studio (common area).
and
Marian Obuseh (marianobuseh@gatech.edu), 
Office hours: Wednesdays, 12 -1pm, Location: ISyE Main 344. 

Recommended Texts: Students are expected to scribe lectures which will be made available. We will post readings specific to various topics and broadly follow the book Introduction to Online Convex Optimization by Elad Hazan. Another helpful reference is the book Prediction, Learning, and Games by Nicolo Cesa-Bianchi and Gabor Lugosi.

Website: http://canvas.gatech.edu/                    

Course Description: At the heart of most machine learning applications today – like advertisement placement, movie recommendation, and node prediction in evolving networks – is an optimization engine trying to provide the best decision with the information observed thus far in time, i.e. the problem of online learning. One must make online, real-time decisions and continuously improve the performance with the arrival of data and feedback from previous decisions. The course aims to provide a foundation for the development of such online methods and for their analysis. At times when these decisions are viewed in the context of the background of the consumers, there is some perceived bias or discrimination. The course will also discuss ethical and legal issues that arise due to such unintended consequences of decision-making and mathematical tools to formalize as well as mitigate such effects.

Outline: The course will have four interleaved components:

  1. fundamental theoretical tools for analyzing online methods, like regret analysis, 
  2. algorithmic techniques for developing computationally efficient methods, like weighted majority algorithm,  
  3. applications to real-world problems like computing the least congested path in a traffic network, and
  4. discussion around ethical and legal issues that might arise due to a perceived bias or discrimination in the decisions.

We will outline important connections to statistics, probability, optimization, graph theory, game theory and policy. We will also have some guest lectures by practitioners from the industry. Since the topic of the course is fairly novel, we will indicate directions of further research and many open questions. This is an advanced class and requires some mathematical maturity.

Credits: 3-0-3

Pre-requisites: 3133 (Optimization) and 2027 (Probability). Prior knowledge of machine learning or regression will be useful but not strictly necessary. We will develop most of the required background in the class, but will expect some mathematical maturity.

Grading policy (all dates are tentative and subject to change): Your graded will be dependent on the performance in the following components:

  • 3 assignments (30%, 10% for each): 10% for every hour that you are late.
  • Scribe a lecture (7%): Tex template will be provided, and students will be expected to edit after one iteration of comments from the instructor. After this the scribed lecture will be made available to class.
  • Class participation (3%): Podcasts, questions/answers, attendance, attention.
  • Midterm exam (30%): October 17
  • Course project (30%): Final presentations in class.
  • Projects can involve survey and summarization of literature review, or an interesting implementation of course concepts to a real-world application, or attempt to progress the state-of-the-art on a novel research question. We will discuss in class and upload potential topics for exploration. Each project team should have 4-5 students.

Course policy

  • All assignments should be submitted online. Every solution has to be in latex format or Pdf. A template will be made available.
  • You are encouraged to discuss homework/assignments problems with your fellow students. But your final answers should be based on your own understanding unless it is a group assignment, which will be announced on Canvas. Copying others’ work is NOT acceptable and violates the honor code. Write the names of students you have discussed the assignments with.
  • Requests for re-grading assignments/exams should be made within a week of returning assignments/exams.
  • The midterm exam will be comprehensive and closed-book.
  • Detailed information about group project can be found on Canvas.

Academic Honor Code: It is your responsibility to get familiar with the Georgia Tech Honor Code and you are bound by its requirements. For any questions any Academic Honor Code issues, please consult me, my teaching assistants, or www.honor.gatech.edu. (Links to an external site.)

Class Presence and Participation. Class presence and participation points are given to encourage your active class participation and discussion. You will be rewarded with a perfect score as long as you frequently come to class and actively contribute to the class discussion during recitations and lectures.

Presence: Although it is not required, most students send their professor a brief e-mail to explain their absence in advance. Students who repeatedly arrive late to the lecture or recitation will have their Class Participation grade lowered. Please sign the attendance sheet when you come to the class. Any false signatures will result in zero participation grades for all parties involved.

Participation: We will devote one entire session to the case discussion. The instructor’s role during a case discussion is that of a moderator. When the cases are discussed, we are less concerned with “right” or “wrong” answers than we are with thoughtful contributions which follow the discussion and either add to the debate or move it in a new direction. If you find it uncomfortable to speak up in class, we encourage you to visit your professor in office hours and work on this skill.