FA19: Deterministic Optimization

This course may be unavailable.

Overview

This course blends optimization theory and computation and its teachings can be applied to modern data analytics, economics, and engineering. Organized across four modules, it takes learners through basic concepts, models, and algorithms in linear optimization, convex optimization, and integer optimization.

The first module of the course is a general overview of key concepts in linear algebra, calculus, and optimization. The second module of the course is on linear optimization, covering modeling techniques with many applications, basic polyhedral theory, simplex method, and duality theory. The third module is on convex conic optimization, which is a significant generalization of linear optimization. The fourth and final module focuses on integer optimization, which augments the previously covered optimization models with the flexibility of integer decision variables.

Syllabus

Week 1

Module 1: Introduction
Module 2: Illustration of the Optimization Problems

Week 2

Module 3: Review of Mathematical Concepts
Module 4: Convexity

Week 3

Module 5: Outcomes of Optimization
Module 6: Optimality Certificates

Week 4

Module 7: Unconstrained Optimization: Derivate Based
Module 8: Unconstrained Optimization: Derivative Free

Week 5

Module 9: Linear Optimization Modeling – Network Flow Problems
Module 10: Linear Optimization Modeling – Electricity Markets

Week 6

Module 11: Linear Optimization Modeling – Decision-Making Under Uncertainty
Module 12: Linear Optimization Modeling – Handling Nonlinearity

Week 7

Module 13: Geometric Aspects of Linear Optimization
Module 14: Algebraic Aspect of Linear Optimization

Midterm

Week 8

Module 15: Simplex Method in a Nutshell
Module 16: Further Development of Simplex Method

Week 9

Module 17: Linear Programming Duality
Module 18: Robust Optimization

Week 10

Module 19: Nonlinear Optimization Modeling – Approximation and Fitting
Module 20: Nonlinear Optimization Modeling – Statistical Estimation

Week 11

Module 21: Convex Conic Programming – Introduction
Module 22: Second-Order Conic Programming – Examples

Week 12

Module 23: Second-Order Conic Programming – Advanced Modeling
Module 24: Semi-definite Programming – Advanced Modeling

Week 13

Module 25: Discrete Optimization: Introduction
Module 26: Discrete Optimization: Modeling with binary variables - 1

Week 14

Module 27: Discrete Optimization: Modeling with binary variables – 2
Module 28: Discrete Optimization: Modeling exercises

Week 15

Module 29: Discrete Optimization: Linear programming relaxation
Module 30: Discrete Optimization: Solution methods

Taught by

Andy Sun

Reviews

3.5 rating, based on 4 Class Central reviews

Start your review of FA19: Deterministic Optimization

Anonymous

This course (run in early 2018) broke the enrollment clause stating "Audit this course for free and have complete access to all the course material, activities, tests, and forums". There was no midterm and final exams for Audit students. Therefore I…

This course (run in early 2018) broke the enrollment clause stating "Audit this course for free and have complete access to all the course material, activities, tests, and forums". There was no midterm and final exams for Audit students. Therefore I was not able to obtain even a symbolic passing grade, despite completing all assignments that were provided to me.
Moreover the staff did not provide responses to the issues reported by the students on the forums. The issues included:

- duplicate video uploads (addressed)
- missing video parts in Week 4 Module 8: Unconstrained Optimization: Derivative Free Lesson 1: Methods for Univariate Functions (not addressed)
- possible typos in the videos (not addressed)
- missing possibility for uploading week 9 and week 10 assignments (not addressed)
- the total contribution of homeworks (25%) that did not correspond to the syllabus listing 40% (not addressed at the time of writing, 2 days before the course closes)
- regular questions asking for clarification of the subject (not addressed)

Sometimes the most grave problems, like wrong video uploads, would get fixed silently, but there was no mention about this fact by the staff on the forums. After weeks 9-11 of the course, when the assignments 9 and 10 appeared without the possibility to submit the homeworks, most of the Audit students left the course. There was a 3 week period without a possibility of submitting homeworks without any response from the staff, so everyone probably just concluded that the course has been forgotten by the staff.

The end results was that there were not enough Audit students in weeks 12-15 to obtain the required 3 peer reviews! I don't know how the course ended for verified students since they stopped participating in the discussions visible to everyone on the forum around the same time.

Multiple attempts of contacting edx support, personally the instructor of the course, and even the board of https://pe.gatech.edu/georgia-tech-online, early in the course have not improved the way this course proceeded.

Now, the presentation and choice of the subjects in the videos were very good. The discussions in the lectures were however mostly intuitive, without much mathematical rigor, overall at a decent engineering college level. The decent academic level given in the lectures is extremely rare in online courses nowadays. Some assignments were challenging for someone unfamiliar with the subject, also due to the practically non-existent practice problems, and due to the absence of the staff on the forum.

Think about this course as similar to "Learning from data" https://work.caltech.edu/telecourse.html wrt. the difficulty level, but without a single helping hint from the staff, and with multiple technical issues.

How any MOOC business developer may think such practices will result in a higher number of verified students, instead of damaging the image of the college?
Stiven

The course is an introduction to deterministic optimization, the topics covered range from non-constrained optimization (the most basic one) to basic algorithms for integer programming, central topics like convexity and applications are also include…

The course is an introduction to deterministic optimization, the topics covered range from non-constrained optimization (the most basic one) to basic algorithms for integer programming, central topics like convexity and applications are also included. Mostly Python is used for solving the problems making used of several open source libraries, MATLAB was briefly mentioned but in my opinion Python is better for this field. The quality of the lectures is very good, and the peer review assessments allow students to have a sort of personalized feedback.

I think this was the first run of the course, so there is room for improvement in a few aspects. The communication between the staff and the students is very poor, we almost never got answer from them, a number of topics in the syllabus were not presented and the order of the topics was changed in some occasions, I was particularly disappointed during the last modules.
SalMo

This course was a basic course to understand linear optimization and methods to reformulate large linear or non-linear models to a solvable/linear model. The course assignments are half theoretical and half computational using Python or MATLAB. I wish it had a second part working more with non-linear models and available tools for nonlinear optimization in Python.

Major problems of the course are lack of STAFF communications on changes/updates and the management of the errors happened by homework submissions and corrections, which were at times nerve racking. Also the syllabus could be more accurate and organized.
Mohamed Chenini

The introductory text says there 4 modules, but the syllabus shows that there are 14 modules.
Please correct or explain the discrepancy.