Ryerson University's MSc in Data Science and Analytics: A Review

13 Sep 2019

I recently finished this program, and sometimes get questions from prospective students. Since it’s not possible to always answer everything in detail to each person, I thought it might be a better idea to write it up in a blog post.

Overall Structure of the Program


This degree requires students to take 4 required courses, 2 electives, 2 seminar courses and a major research project. For full-time students, the courseload usually looks like:

Fall -> 3 required courses

Winter -> 1 required course + 2 electives + 1 seminar course

Summer -> 1 elective course + major research project

So technically, you can start the program in September and finish it by August next year.

Courses


DS8001 - Design of Algorithms and Programming for Massive Data

This was taught by Dr Andriy Miranskyy. I found him to be an excellent professor who had subject matter expertise, genuinely cared about the course quality and was willing to listen to feedback, and this was easily one of the best taught courses in the program. It’s a required course which teaches you about algorithms and their applications for big data. The project component in the course helped enforce the knowledge taught in the course. My project for this course, for example, is here

DS8002 - Machine Learning

Another required course, which was taught by Dr Ayse Bener. The course format was exams and lab exams, where you have to code machine learning algorithms from scratch in a given time frame.

DS8003 - Management of Big Data and Big Data Tools

This was a more application-based required course, taught by a professor who has since left. The format was all take-home assignments and exams. I found this to be the course that required the least effort in the first semester, but the knowledge is useful for the industry.

DS8004 - Data Mining and Prescriptive Analysis

Nope. This course was a disaster in terms of quality, based on how it was taught and graded. If I could, I would advise not to take it, except you can’t, because it’s a required course. It’s essentialy machine learning part 2, except I would learn better if I took an online course instead.

DS8007 - Advanced Data Visualization (elective)

Disclaimer: I did not take this course

Opinion: It’s data visualization and I didn’t think I needed a whole course in a Master’s degree to learn about data visualization, I could probably do that in a weekend online. Besides, I already knew quite a bit about HTML, Javascript and D3.js which they were teaching in the course. I would classify it as a bird course.

DS8008 - NLP (Text Mining) (elective)

This was taught by Dr Shariyar Mortaza. It was well-taught and rigorous, and gave a good foundation of NLP. The format was lab exercises, course project and a final exam quiz. For an example of what sort of project this course might require you to do, you can take a look at my project here

DS8009 - Special Topics in Data Science and Analytics (elective)

In the Winter semester, the topic of choice was Deep Learning

Disclaimer: I did not take this course

Opinion: I didn’t need to take a whole course on Deep Learning since so many resources on it are readily available online, and it’s a fast-moving field so by the time the course is finished the knowledge taight would probably already be outdated. I leaned more towards taking courses that I would get more value from.

DS8010 - Interactive Learning in Decision Process

Forget the obscure title: this is “Reinforcement Learning”.

Disclaimer: I did not take this course

Opinion: I have heard that the course was well-taught and useful. I did not take it simply because I was not interested in the particular topic.

DS8011 - Bayesian Statistics and Machine Learning

This was taught by Dr Dhanya Jothimani. It was quite theoretical, which was what I was looking for since I am trying to develop my theoretical knowledge in machine learning. The course provided a primer on the Bayesian perspective, and I thoroughly enjoyed it because I had more leeway to ask questions and get more out of it because it was a small class and the course was well-instructed. I have since become interested in further studying Bayesian machine learning and probabilistic programming. Thanks Dhanya !

DS8005 - Soft Skills, Communication and Ethics

and

DS8012 - Research Skills

Seminar courses. Essentially you don’t really have to study anything for these courses, you just have to have a particular level of attendance and make a presentation.

Major Research Project


I did my major research project with Dr Naimul Khan. He was helpful and provided good, prompt feedback in terms of the project. The topic was “Developing a Confidence Measure Based Evaluation Metric for Breast Cancer Screening Using Bayesian Neural Networks”, and it was a useful component of the program. You can view my MRP here. At the end of the project you have to attend a poster presentation, where you present your work to your supervisor and second reader in order to get the MRP milestone approved.

Students


I would say that only a small percentage of students in the program (or at least, in my batch) were serious about actually learning. The rest were just there to get the degree, and to somehow get into “Data Science”. I noticed several students cheating in the lab exams, assignments and carrying out severely non-ethical violations; on occasions I had to work with students who wanted a free ride in terms of their share of work in group projects, wanted more credit in projects despite doing very little work, blatantly tried to copy my work/steal credit. In the end they got away and will graduate with the same degree as me, which is disappointing.

Common Questions I Get Asked


What are the job prospects of this program?

Opinion: This program will get you started with machine learning. You have to do a lot of studying and work on your own. You should not expect to land a data scientist role just out of this program. Based on what I have seen of the industry, the strongest technical roles as Data Scientists usually require PhDs or very strong work experience. However, the degree can get you started in that direction, and you can slowly move up, provided you work hard.

Fact: I am currently on a break till early 2020 and will update this post once I start interviewing for a Machine Learning/Data Science role.

How much do you get paid after completing this program?

There’s no short answer, it depends on what sort of role you land. Take a look at standard salaries on Paysa for the role you want, to get a better idea of this.

What are the prospects for research in this program?

This is not a thesis-based program. It’s an applied program with 1-year of coursework + project. There may be some opportunities for research if you get into the Data Science Lab, but sometimes the people you work with in the lab may not be worth it. There may also be some opportunities for research based on who your supervisor is for the Major Research Project. In the end however, the research component is not a focus in the program and the program is upfront about it.

Other notes of interest


  • Some instructors seemed to take it personally when you didn’t take their courses, which is fascinating considering the fact that you would expect postdocs/lecturers to behave in a mature way
  • I am thankful to Dr Ayse Bener, who is the Program Director. She listened to my feedback, and was very helpful in ironing out issues I had with the instruction and grading of a course. I found her to be a fair and impartial administrator. If you have legitimate issues with anything in the program, please contact her directly.
Web Statistics