Ms data science for 15k

[dutchdragon]

You brought up an excellent point. Five grand is five grand, which is a lot of money. In fact, let me tell you what I did. I wanted to switch to a machine learning career. I originally enrolled in the MSDS program at Eastern. After completing six courses I was really disappointed. The courses were just too easy (100% in every course), and I wasn't learning anything. So then I questioned whether I wanted to spend another $4k there to complete the degree or use the money to help pay for the OMSCS program at GT instead, which is what I did.

I still want to do a Data Science Master's, so I started the Data Structures pathway at CU, but the first course was way too demanding (be careful what you wish for, right?). It's like they jammed in 3 units worth of work in 1 unit. I dropped it and instead switched to the Statistical Inference Pathway.

I also considered the Master of Data Science at Illinois Tech on Coursera. If you do some Coursera Specializations, you can get credit for up to 6 units and get the Master's for about $12,300. Performance-based, no application, no transcripts, just need Bachelor's. But with CU, the 1-unit granularity, being an AAU school, and more recognized here out West (I'm in Cali), I'm going to attempt the CU program first.

Thanks for sharing your experience. You have given me much to think about. (I actually already have those coursera certs to transfer to Illinois).

So an update in my progress on the CU Boulder program. Had to drop the course    . The probability course was just as demanding as the data structures. It felt like I was doing a 3-unit course for just one unit of credit. So now if I finish the program, it's going to be almost $17k for me as I'm out $1,050 now. Also, if every course is like this, then I just don't see myself doing 90 units of effort to earn 30 units. But I don't know, maybe I'm the problem    . I checked online to see if anyone else did this program, and apparently one person has a Medium post back in 2021 saying he started the program. I then Googled his name and checked his LinkedIn and looks he completed the degree in 2023, so it does look doable.  And now he has a nice position in data science.  I don't know, I might look elsewhere. I won't go back to Eastern. If a dumb person like me says Eastern was too easy, that tells a lot about the program.

@ ifomonay Yikes, you're having me worried now. I'm still going back and forth deciding between Eastern and UC Boulder. 
To put the difficulties you're having with Boulder in perspective, do you come from a STEM background or not? (I'm sorry to hear you had to drop the course.  )

Sorry, I missed this! Okay, so based on my experience, I think Eastern is better. Even though it looked like I trashed the program, there are some positives, which I think makes the learning experience good:
1. The administrative staff is excellent. If you have registration issues, etc, they will help quickly and not in a condescending way.
2. The courses do only scratch the surface and is too easy, but, if you have discipline you can get way more out of the courses. The GA's (Graduate TA's) they pick are industry experts. If you go through the optional or supplemental readings and ask questions, they will answer in the course discussion board. So, in short, easy A with little effort, but great opportunity to learn a lot if you can somehow do more in the 7 weeks.
3. There's a Discord which is very active, and fellow classmates are very supportive. You can't discuss hw problems, but you can ask lecture questions.
4. Also, to be fair, the program was never meant to be heavy on theory anyway and is meant to be a professional masters (despite being a Master of Science). It was intended to be for those without a quantitative background, and really geared to those who want practical skills that employers are looking for today. I left the program about 2 years ago, and a lot can change over 12 terms, so my experience may also be outdated.

About Boulder, I won't continue the program.
1. It seems to me that it's oriented towards students without bachelors degrees. It seems like they're saying, "since you didn't spend four years doing a BA/BS, we're gonna make you spend that time in this program". I don't want to do 90-units of work to get just 30 units. That's just my feeling though. Maybe others here who've done he MSDS program at Boulder can give a recommendation, but I won't.
2. Overall it feels lonelier. You feel like you're the only one in the course. You do get anonymous peer reviews, but it's just check mark boxes on whether you did the problem or not. There is a Slack, and most students are friendly, but there a lot of smart alecs there too. Like if you ask ,"I still don't know the difference between conditional probability and Bayes Theorem". There will always be someone who'll answer, "watch the lecture videos".
And yes I do have a STEM background. EE and CS (OMSCS). Previous degrees were challenging but taught/presented in interesting ways. The lectures at Boulder are so boring. Just presenting pages of formulas because they have to, and then a lot of problems. Many concepts to do the problems are not in the lectures, but are in the textbook.

Thank you so much for still responding to this. 

I have decided (at least for now) to go with Eastern University for a couple of reasons.

1.  It has a "small uni feel" which I like, and the instructors (especially Greg) seem amazingly approachable people from the video chats I've seen. I'm glad that you confirmed my suspicion that this is a stark contrast to Boulder, which feels "lonelier" as you say. This was indeed the vibe I got from the pathway courses I did.

2. Even though I have a CS background, and in that aspect, the program is not 100% targeted to me, I do feel like with the supplemental material I will be able to self-teach and dive deeper into things that interest me should I want to. I'd rather be taught the absolute "need to know", with the option to dive deep, rather than, like you say is the case with Boulder, having to take 3 credits worth of stuff for each credit earned. Especially if most of the math isn't used in a practical manner for most data science jobs anyway. If I were eyeing a PhD in the future I might have considered it.

3. The biggest pro for me, is that I can finish this program fast. I'm in my mid-40s and only just about to finish my Bs in CS. While I've really enjoyed going back to school mid-career, I'm looking forward to finishing my masters within a year and getting it over with.

Now that you've stopped Boulder, are you eyeing a different master's program?

I decided to pause on doing an MSDS altogether. There seems to be a new one popping up everyday, so maybe in a couple years I'll look into an MSDS program again. Instead, I applied to and was accepted to MS in Applied Statistics at Purdue. It seems like probability and stats are very useful for AI and ML (plus the Boulder experience made me realize I'm weak on the statistics side). Some courses require SAS (doesn't sound interesting, oh well), but I think most other courses use R.
Ok, for you, I think Eastern is a great choice for a quick MS degree to get that out of the way. With a BSCS, I think it would be criminal if you didn't go on and do the OMSCS at GT, so add that next. It's less that $8k, but I don't think it's that quick, maybe 2-3 years. I used the full 6 years.

I would've considered GT, but (for now) my Bachelor's is not regionally accredited, which is a requirement there.

Good luck at Purdue! Applied statistics is a great choice.