Overview | Tharindu Cyril Weerasooriya

Evaluation

Sun, 05 May 2019 00:00:00 +0100

Grade Categories

Category	Contribution
Homework Assignments (3)	30%
Class participation	5%
Project Proposal (1)	15%
Project Milestone (1)	20%
Final Exam	30%
Total	100%

Numerical grades will be converted to letter grades according to the following scale: 90%: A; 80%-90%: B; 70%-80%: C; 65%-70%: D; < 65%: F.

The lowest graded homework/project will not be dropped.

To calculate final grades, for each of the graded categories above (i.e., final exam) we will take the raw score given in the Grade tab and scale it to the points give above. E.g., if G is your final exam grade, and it is out of 40 raw points, the total number of final points is 0.25 * (G/40).

Homeworks

There will be three homeworks. They will be made available in the Contents section, and you will be notified via the News widget on MyCourses.

You are not allowed to discuss assignments with anyone except CS tutors (but not the grader) and me. You are also not allowed to look up the answers to your assignments or at other people’s work, eavesdrop on other’s homework discussions, etc. You should submit only work that is completely your own and you should be able to explain all of your work to me.

Submit your assignments by uploading it to the appropriate MyCourses dropbox, as a tar.gz file.

All assignments must contain a pdf (Word documents, text files, etc. will not be accepted nor graded) containing all answers to all problems, except for coding problems. For coding problems, the pdf document must contain the full instructions for running your code and an example for how to run it from the command line.

All code must be written in Python 3, unless otherwise noted. You are strongly encouraged to use the Anaconda distribution of Python.

I will stop answering homework questions the day before they are due.

Projects

There will be two projects. These must be done in groups of 3-4. The same rules regarding sharing information apply here, except of course that may share and collaborate freely with your partner(s).

For both projects and homeworks your include your name (and, for projects, your partners’ name(s)) on all documents submitted (including source code). Source code should have comments at the beginning of each function describe what it does and describe each input and output.

Generic Rubric

When possible, I’ll post more specific rubrics, but this gives a sense of how the points allotted to each question will be allocated.

100-90%	90-75%	50-75%	<50
Work complete and mostly error-free. Writing at professional level.	Work mostly complete, but contains errors. Writing not at professional level.	Quality or work inadequate, but evidence of effort. Sloppy writing. Code doesn’t work.	Work incomplete, little evidence of effort.

Disputing Your Grade

If you feel that an error was made in grading your assignment or exam, you have one week from the moment the graded work was handed back to dispute your grade. All grading issues should be taken up with me; do not discuss grading issues with graders or tutors!

To submit a work for regrade, send me an email with REGRADE [assignment name] as the topic (e.g., “REGRADE homework 3”). To help expedite the process, please attach all relevant material (such as the grader feedback). Then explain in the email why you think you deserve a regrade. You may refer to the particular problem(s) in question and even provide in the grader feedback (e.g., by annotating the pdf if the feedback is itself an annotated version of your work, which is often the case) additional details where the problem appears.

Academic Dishonesty

You should only submit work that is completely your own. Failure to do so counts as academic dishonesty and so does being the source of such work. Submitting work that is in large part not completely your own work is a flagrant violation of basic ethical behavior and will minimally be punished with failing the course.

Syllabus

Sun, 05 May 2019 00:00:00 +0100

Textbooks

We will assign readings from the following texts.

by Goodfellow et al.
by Kevin Murphy

Other Supplemental Texts

, by Tom Mitchell. Great introduction to machine learning

Web Resources

Theoretical Computer Science Cheat Sheet Definitions Series has some useful formulae for this class, but also for others, including algorithms.

Schedule

Week	Start Date	Topic to be covered	Project Milestones	Homework
1	5/14/2020	Introduction, data, real world issues, linear algebra, and statistics
2	5/21/2020	Numerical methods, optimization, calculus		HW 1 \[Release\]
3	5/28/2020	Supervised Machine learning
4	6/4/2020	Feedforward neural networks, Keras, Tensorflow	Project 1 due	HW 1 due
5	6/11/2020	CNNs		HW 2 \[Release\]
6	6/18/2020	Regularization over neural networks and other practical topics
7	6/25/2020	Optimization		HW 2 due
8	7/2/2020	Sequence learning: Markov models
9	7/9/2020	HMMs, LSTMs	Project 2 due
10	7/16/2020	Dimensionality reduction		HW 3 \[Release\]
11	7/23/2020	Unsupervised Learning
12	7/30/2020	Reinforcement learning		HW 3 due
13	8/5/2020	Wrap up	Final Exam \[Aug 9\]

Tech Logistics

Sun, 05 May 2019 00:00:00 +0100

Experiments for Homeworks & Projects

I expect you to be familiar with writing code in Python 3, and . I expect you to be familiar with running a notebook and answering the questions related to a homework using it. You do not need to own a machine with a GPU, but it does make things faster (TBA: Access to a CS Server with a GPU which you can use for completing Homeworks and Projects).

- Google provides free tools to learn machine learning and Colab is its own Jupyter notebook server. Consider it Google Docs, but for ML. Best of all, you do not need to configure anything, it is ready to run out of the box.

- Another great solution from Google for their Cloud computing, which lets you use their GPU VMs for a 1 year without any cost.

Project Reports

For project submissions, the code should be accompanied by a properly formatted report using LaTex. I recommend using for your collaboration between team members.

The submission should be using the .

Introduction

Sun, 05 May 2019 00:00:00 +0100

Welcome to the first day of class! We would be having our first class as a pre-recorded session, but based on your responses (time zone mostly), we’ll soon be meeting live via Zoom.

The grading rubric was updated from the video to add a contribution for participation

Resources Mentioned in this lesson

CS Common Course Policies

Sun, 05 May 2019 00:00:00 +0100

Rescheduling an Exam

Exams can not be made up except for real emergencies in which case proper documentation (like a doctor’s note) will be required. If at all possible, you should contact me prior to the exam. Oversleeping, cars that don’t start etc. do not constitute a valid excuse.

Please refer to the for related questions.

Course Withdrawal

During the add/drop period, you may drop this course and it will disappear from your transcript. After that time, you can only withdraw from the course; the course will appear on your transcript with a grade of W. See the institute’s regarding the add/drop period and latest withdrawal date.

Disability Services

RIT is committed to providing reasonable accommodations to students with disabilities. If you would like to request accommodations such as special seating or testing modifications due to a disability, please contact the Disability Services Office. It is located in the Student Alumni Union, Room 1150; the web site is . After you receive accommodation approval, it is imperative that you see me during office hours so that we can work out whatever arrangement is necessary.

Academic Integrity

The will be enforced.
You should only submit work that is completely your own. Failure to do so counts as academic dishonesty and so does being the source of such work. Submitting work that is in large part not completely your own work is a flagrant violation of basic ethical behavior and will be punished according to department policy.

Mon, 01 Jan 0001 00:00:00 +0000