ECE 595: Homework
Professor Stanley H. Chan
Spring 2019
Submitting Your Homework
Due Date / Time
Homework is due 4:30pm on the due day. Please print a hardcopy of your homework, and put it in the dropbox located at MSEE 330.
Late homework will be penalized according to the following rubric:
Same day, by 5pm, 20% (1 point) off. This includes printer failure and reasons like that.
Next business day, by 4:30pm, 40% (2 points) off
Afterwards, we do not accept.
If you need to travel on the homework due day, you need a written approval from the instructor before the dealine.
Class ID
Given the large number of students of this class, we kindly ask you put your
Class ID on your homework. This will help us tremendously sorting your
homework. Thank you!
Homework
We encourage you to typeset your homework using LaTeX. For instructions, please visit the LaTeX page.
Homework 0 (PDF) (Do not handin)
Homework 0 is a prerequisite check. Please try and see how comfortable you are.
If you can comforably do three or more problems, you should be fine. By comfortably solving the problem, we meant that you know the mainstream approach and are able to apply it. (Mainstream approach = Something you would normally see in a textbook. No hacking.)
If you can only do two problems, you will probably have some difficult time in this course.
If you can at most do one problem, we encourage you take the course at a later time.
Homework 1 (PDF) (Due: Jan 18, 2019 Friday, 4:30pm)
Solution
Supplementary Note (PDF)
Homework 1 focuses on linear algebra and probability.
Please read Chapter 1 for review, and references suggested in the chapter.
You will be asked to write Python codes to generate plots. Please embed your code into your homework report, with write ups.
We encourage you use the LaTeX template when typing your homework report.
Homework 2 (PDF), data
(ZIP) (Due: Feb 1, 2019 Friday, 4:30pm)
Solution
Supplementary Note (PDF)
Homework 2 focuses on optimization.
Please read Chapter 1.3, and references suggested in the reading section of this website.
We need to use CVXPY for this project. If you need help, please reach out to one of our TAs.
As usual, we encourage you use the LaTeX template when typing your homework report.
Homework 3 (PDF), data
(ZIP) (Due: Feb 22, 2019 Friday, 4:30pm)
Solution
Homework 3 focuses on Bayesian decision rule.
Please read Chapter 2.1 and 2.2.
You may find Duda Hart Stork chapter 2 and chapter 3 useful for this project.
If you need additional help on the programming, you can check out ECE 302 Project 2.
As usual, we encourage you use the LaTeX template when typing your homework report.
Homework 4 (PDF) data
(ZIP) (Due: March 22, 2019 Friday,
4:30pm)
Solution
Homework 4 focuses on logistic regression, perceptron algorithm and SVM
Please read course note chapter 2.3  2.6
As usual, we encourage you use the LaTeX template when typing your homework report.
Homework 5 (PDF), data
(ZIP) (Due: April 5, 2019 Friday, 4:30pm)
Solution
Homework 5 focuses on adversarial attack
Training data is the same as homework 3
Please read course note chapter 3
If you need additional help on the programming, you can check out ECE 302 Project 2.
As usual, we encourage you use the LaTeX template when typing your homework report.
Homework 6 (PDF), (Due: April 26, 2019 Friday,
4:30pm)
Solution
Homework 6 focuses on learning theory
Read course note chapter 4
Read Learning from Data chapter 2 and chapter 4
If you have any question, please contact one of the TAs
As usual, we encourage you use the LaTeX template when typing your homework report.
Grading
Homework will be graded on a coarse basis. For each homework, you will receive:
5 points. All correct / small typos.
4 points. Minor mistakes.
3 points. Some mistakes.
2 points. Major mistakes.
1 points. You hand in, but none of them is correct.
0 points. You do not hand in homework.
Picking up Graded Homework
Graded homework will be placed at the secretary's office at MSEE 330. However, per FERPA policy, we need your permision in order to leave the homework unattended. Please fill up the following form to tell us whether you give us the permission:
Homework Return Policy (Website)
There are two options.
YES. (Default Option) I agree to leave my graded homework unattended in the secretary's office in MSEE 330. I understand that other students may see my score, but I do not mind.
NO. I like to pick up my homework from the secretary in person. I understand the inconvenience, but I want to keep my score confidential.
