Eastern University MS in Data Science 2022 Review

Introduction

I recently finished the MS in Data Science at Eastern University. This is a 30-credit program that is designed to be taken part-time and finished in 10 or 20 months (depending on whether you take 2 courses or 1 at a time.) The program uses accelerated 7-week courses and has 6 terms per year so that you can get trained as a data scientist in a hurry.

I started in January 2021 and I’m slated to graduate in May 2022, pending a passing grade on my capstone. (Update: I passed!) I previously made a post when I first started and another one when I got about half way. This post summarizes my final thoughts of the program now that I’m finished.

The courses that I had not completed at my half-way review included DTSC-600, DTSC-670, DTSC-680, DTSC-690, and DTSC-691.

DTSC-600 Information Visualization

This course I actually took near the end, along with 690. This course has no coding, but instead focuses on Qlik and Tableau. There are two required textbooks:

• Information Dashboard Design by Stephen Few. 2nd Edition. Publisher: Analytics Press
• Storytelling with Data: A Data Visualization Guide for Business Professionals. Publisher : Wiley; 1st Edition

For each module, you’ll do some reading and complete a quiz. Then you’ll create a dashboard using a given dataset (e.g. Australian Tax Revenue or Console Sales) and then answer a second quiz made up of questions based on the dataset.

This course is being retooled, so it will actually include more matplotlib in the future. I can’t comment on what that might look like going forward. I didn’t find this course super helpful because I already have a background in Tableau.

DTSC-670 Foundations of Machine Learning

This was a super fun course, though challenging. There are 9 assignments. The assignments involve doing data cleaning, building machine learning models, building pipelines, and other things that will be useful for the capstone. It uses Python. You can read my DTSC-670 Tips and Tricks article for more information about this course.

DTSC-680 Applied Machine Learning

This is a continuation of 670. A very challenging course, but also really helpful to improve your matplotlib graphing skills and machine learning knowledge. You build a KNN classifier in this course without actually using the KNNClassifier. You also do a complete end-to-end machine learning project without any scaffolding code to guide you.

You do some matrix math (that wasn’t too bad) and you also build an artificial neural network (ANN) using TensorFlow. I haven’t spent much time with either TensorFlow or PyTorch (which is becoming more popular) so this was a heavy lift. I still need to re-watch the 3 hours of video on ANNs in 680 and make sure that I brush up on my knowledge.

DTSC-690 Ethics in Data Science

This was a nice break from all of the coding in 670 and 680. There are 40 readings throughout the term. The median length was 6 but they range from 4 to 8.

For each set of readings, you write one substantive post (I think 250 or 300 words was mentioned in the syllabus) that synthesizes your thoughts for the whole week’s readings and post it on a discussion board. Then you reply to at least 3 students reflections with a similar 250 word post and do that for the 7 weeks.

That makes up about half the grade. The other half of the grade is a presentation on a topic of moral, ethical, or religious significance. My presentation was titled Machine Learning, National Security and Civil Liberties: Technology in Conflict.

Each student gets placed in a group (mine had 8 students not including me) and you need to watch each presentation and give feedback on it, similar to the 250 word posts. The presentation was due part-way through the term, so that you have time to give feedback.

DTSC-691 Capstone

The big fish! This is what all of the program is leading up to. Your capstone. There are three options for doing the capstone right now:

Standard Machine Learning Project

This is what most people think of, when they think of the capstone. You collect some data, complete exploratory data analysis, build a machine learning model, evaluate it, and then create a way to interact with your model.

Database Project

In the database project, you use concepts similar to DTSC-660 where you build a database, create an ERD and schema, and then write functions in Python to work with that database, and create a way to interact with it.

Data Science in Education

In this project, you build a Jupyter Notebook and at least an hour’s worth of lecture to go with to teach a concept in data science education.

Proposal

Early in the term (or even before the term starts) you’ll choose which project you want to go with. Once the term starts, you’re assigned a project mentor. This is a previous graduate or professor who will review your project proposal and provide detailed feedback. While they can’t provide specific help with debugging or coding like a GA would, they can provide you with advice.

You can see an example proposal here. This is very similar to the one I submitted. My mentor provided direction, asked questions and helped me improve my actual proposal, which was on a different topic.

After the Proposal

After the proposal is approved, you get to work! You have the rest of the term to complete your project. I didn’t talk to my mentor much throughout the term, but that’s because I felt like I had a strong proposal and so I didn’t need much clarification.

One thing that is important is that your proposal must provide a way for you to interact with your model outside of Jupyter/Spyder. This means using something like Flask. Flask is its own mini-language that took some learning so I would recommend practicing with it before you need to use it in 691.

Submitting the Final Project

The final project consists of three components: your completed code, a project submission (which is similar to your proposal but instead of showing what you would like to do, it shows what you actually did), and then a 30 minute-maximum walkthrough video that goes over your project.

My mentor gave me detailed feedback on my Python. I was really surprised and impressed. He walked through my code closely and made many recommendations. I hope I can be like him when/if I’m ever a mentor in the future!

Now I’m just waiting for my final grade. This is a Pass/Fail course so I’m hoping that I passed and then I’ll graduate sometime next month. Edit: I got an email letting me know I passed, which is the last hurdle. Now all that’s left is to officially receive my diploma.

Final Impressions

This program is definitely one where you get out of it what you put into it. I’ve learned so much over the last 16 months and look forward to continuing to build my data science skills. For the price, I think the program is really great. You could complete a bootcamp for $10K and come out of it with the same skills but no credential.

I think students should try to build and maintain a portfolio throughout the program which will increase their ease of job hunting when they graduate. For what it’s worth, I got a job with a significant raise (and Data Analyst in the title) while I was in the program and I attribute that to Eastern.

One more thing: the program continues to change. Courses are being updated each term. For example, DTSC-575 and DTSC-650 are already different from the versions I took, even though I only took them 6 months ago.

They are rolling out new courses and an updated course plan where you only need to take 4 courses required: 650, 660, 670, and 690. Then you’ll take 6 electives. Right now there is one elective being tested now and several more planned.

Lots of exciting things. Good luck everyone!