MS in Data Science: self-paced, 10 months, under $10k

Discussion in 'IT and Computer-Related Degrees' started by Seylan, Jun 5, 2020.

Loading...
  1. Dustin

    Dustin Well-Known Member

    Yes. The final is 10% of the grade, the exams are 60%, and the CodeGrade assignments are 30%. You can repeat the 30% CodeGrade and the 60% exams until you get the grade you want.

    If you get 100% on the CodeGrade assignments and 80% on the tests, you'll have an 86 going into the final project (30 + 48 / 90 = 86.67%), so you would only need to get 20% on the final to finish the course with an 80 (30 + 48 + 2 / 100 = 80%). You can technically get below an 80 in the course and still graduate, because it's your cumulative average that needs to be above 80, across all the courses. They try to get everyone to 80% in every course though, to make sure you're absorbing the material.
     
  2. Unistic

    Unistic New Member

    Thanks again for the response, really hoping I get accepted lol. I know its a relatively new program. I did apply Sunday hoping to hear back on Tuesday. I do have an IT bachelors degree but retired military and I dont have to work so no resume (they asked for one of course). I just want to learn, get my master degree and used up the remaining of my GI Bill, plus the degree seem geared more to techniques vs constantly writing papers getting tired of having to paraphrase the same things over again while my pc does all the APA formatting for me. So never felt like I can truly learn when all i'm doing is writing and not applying. Currently taking an hour break before getting back into this YouTube series on Python, planning on doing 8 hours of it a day for two weeks and make a few projects on Github before moving on to R and ML in hopes of being accepted into this program. Again thank you!
     
    Dustin likes this.
  3. Dustin

    Dustin Well-Known Member

    You'll be more prepared than most people with that preparation. I completed the Data Analyst with Python track on DataCamp, which is currently 62 hours of material (I did it a couple years ago so not sure if that's still accurate), and I think it was good preparation for 520 and probably 575 as well (the second Python programming course.)
     
    Unistic likes this.
  4. JoshD

    JoshD Well-Known Member

    Coding them is not too bad. Interpreting them is the difficult part.

    I have no idea how to even explain interpreting them on here to be honest. Took some work on understanding what the coefficients meant for a linear regression and what the exponentiated coefficients meant for a logistic regression.
     
    Dustin likes this.
  5. Dustin

    Dustin Well-Known Member

    I'm going to have to keep watching the videos and reading the book until it "clicks." I know that the unstandardized logistic regression coefficients represent the amount of change in the dependent variable from a one unit change in the independent variable, but I still can't just look at the output of the model and know what it means immediately. We also learned to use the AIC to compare models (lower AIC is better, but can only be used to compare variations of the same models), and using deviance residuals to assess the fit of a model (lower is better there too.)

    The final project is pretty cool. The dataset is 515MB. 441,457 rows and 330 columns: https://www.kaggle.com/cdc/behavioral-risk-factor-surveillance-system

    It's going to take a while. The course ends June 27 so I've got lots of time to work on it.
     
  6. Unistic

    Unistic New Member

    (another message from the guy hoping to get accepted)
    May I ask how far into this program is this module you're referring too? And I'm guessing due to the program being new there's not alot of tutoring assistance offered from the school yet? But myself I'm okay with daunting Youtube videos going over the basics of what I want to learn. But could you also use outside paid tutoring if you choose? Because as I understand it you need atleast a minimal grade in a module to move forward not meeting it within that 7 week deadline is a fail. Just curious, thank you! Hoping to hear from them today.
     
  7. Dustin

    Dustin Well-Known Member

    This is the third course. The first course is Intro to Data Science and focuses on basic Python. Second course is Intro to Statistics and focuses on basic R. Third course is Data Analytics with R and it's a more involved R course. Linear regression is touched on in the second course. Linear and logistic regression come up near the end of this course. A linear regression predicts a Y value from an X value, where both are numbers (e.g. hours of exam preparation predicting grade on the exam.) Logistic regression predicts a Y value from an X value, where one of them is categorical (e.g. exam prep Low/Medium/High or Gender Male/Female predicting grade on the exam.)

    As for tutoring, in addition to the videos and textbook that make up the courses themselves, there are Discussion Boards monitored by Graduate Assistants (GA). The GAs are all students who completed the course with an A and have a pattern of helpful responses. In the lower courses they answer questions on the discussion boards and in the upper courses they also do some grading. The professors also organize live Zoom meetings (that are recorded) at least once but usually twice per term, to ensure you can get questions answered.

    There's also a data science team email for specific questions if the GA isn't sure, and the professors and other staff provide expert assistance there. You could use paid tutoring if you wanted, though I don't know that you'll need it. And then there's a fairly active Discord, which is a chat platform open to students and alumni.

    You need that 80% minimum on the exams but you can earn less than an 80 in the course as a whole and still move on to the next course. You need a cumulative 80% (3.0) to graduate. If you find yourself unable to complete the course requirements due to emergency circumstances, you can request an "Incomplete" and if approved you can be granted up to 180 days to finish the course requirements before your grade converts to an F. https://www.eastern.edu/sites/default/files/sites/default/files/offices-centers/registar/incomplete_grade_agreement.pdf
     
  8. Unistic

    Unistic New Member

    Thank you again for responding. The way you explain it make it seem really well put together! Crossing my fingers in hopes of being accepted and starting as soon as possible.
     
  9. Unistic

    Unistic New Member

    I got approved!!!

    Like I mentioned I only have 16 or so months of the GI Bill remaining the degree can be finished in 10 months which will leave me with a few months to use on a boot camp or certificates if I see the need. But of course to complete it within 10 months I'll have to double up. And from what you're saying its not always a good idea. Some ppl I have messaged on the FB group have been doing 2 courses off and on as well. The advisor said if I want to register two to start I'll have to do:

    Course Number: 520
    Title: Fundamentals of Data Science
    Course Number: 550
    Title: Intro Statistical Modeling

    What are you thoughts? Even though I have some experience I consider myself an absolute zero beginner. Should I take both courses at once? Time wise I'm home 24 hours a day since I'm retired and taking care of my kids. I can devote 8 to 12 hours a day to the course throughout the day. I still don't want to get stuck, lost etc. Even though the advisor also said dropping or picking up a course will not be a problem. I'll like to hear from you since you have recent experience. Thanks again!
     
    Mac Juli likes this.
  10. Dustin

    Dustin Well-Known Member

    Congrats!!

    520 and 550 are both exam based, and mostly theory. A bit of Python syntax. I think you'd be fine to do 520/550 at the same time, especially given that you have more free time. 520, 550, 575, 600 are the fastest/simplest courses. You don't take them in that order, but those are ones I would consider doubling up.
     
  11. Unistic

    Unistic New Member

    Thank you again!
     
  12. JoshD

    JoshD Well-Known Member

    Dustin,

    I am curious how the Program Director got his PhD in Psychology and was a Psychology Professor but then became an Assistant Professor of Data Science at Eastern? Did he have an undergraduate or masters degree to aid in that? If not, I need to get with him because moving into the Data Science world is TOUGH. I can not see how one becomes a Professor on the topic without an extenuating amount of experience and/or education.
     
  13. smartdegree

    smartdegree Active Member

    Data science is a relatively new field and data science degrees really only started to show up the past 5 years. So makes sense I guess that the most established data scientists don't have data science degrees? Psychologists have strong training in statistical methods which makes them good candidates I guess. Same with CS / stats / economics grads.

    I have a masters degree in econometrics and to be honest, machine learning models are actually pretty straightforward to understand and use compared to econometrics. There is a lot of debate currently within economics on whether to adopt machine learning but the traditionalists are either too lazy to learn new stuff or see machine learning as too radical for their tastes. I feel it's inevitable though that economics ultimately moves towards machine learning.
     
    JoshD likes this.
  14. Dustin

    Dustin Well-Known Member

    Yes, the Program Director has his doctorate in Experimental Psychology. The other professor has his doctorate in Computer Science. There's a third Professor who teaches a class in Data Visualization, I'm not sure of his specific credentials. (His undergrad and Masters were presumably also in Psych.)

    Data science is an interdisciplinary field (a blend of computer programming, applied statistics and domain knowledge), so there are a very limited number of data science PhD programs and Professors tend to come with PhDs from other disciplines like Computer Science, Economics, etc.

    Prof. Longo teaches the intro and applied statistics courses based on his background, Prof. Morabito teaches the computer programming (including the machine learning courses) and then we bring into the field our domain knowledge.
     
    JoshD likes this.
  15. JoshD

    JoshD Well-Known Member

    Absolutely! My professors for my data science oriented courses at Duke have their Masters and/or PhD in Statistics, which makes complete sense. I was just curious how a Psychologist wound up as an Assistant Professor of Data Science as I would not have guessed data science being a part of a Psychology PhD program.
     
    smartdegree likes this.
  16. Unistic

    Unistic New Member

    I get what you're saying and I understand that as well. But the masses will only see that its a Data Science Master Degree, Regionally Accredited, and Under 10k, the accreditation and price is the main selling point. I have friends whit almost over 50k in student loan debt and they can't do anything but work and squeeze every penny to pay that off and its crazy. You can basically pay out of pocket with this degree. In this field/programing/DS you can self teach yourself using only resources/books. I see this is a check mark for higher paying for those who are just doing it and cheap learning experience for those who need it. Supposedly they're doing a 10 month MBA at the same price....Might have to hop on that one as well.
     
  17. Dustin

    Dustin Well-Known Member

    Completed DTSC-650, Data Analytics with R. Before the final project grading I have a 95.8% (86.22 out of 90 points.) I need to get 68% on the final project to finish with a 93. The final uses this 515MB dataset: https://www.kaggle.com/cdc/behavioral-risk-factor-surveillance-system

    You can still submit until the last day of the course. I refactored my R code a bunch, but the only thing I'm really not confident about is the quality of my visualizations. ggplot is definitely different than matplotlib.
     
  18. Dustin

    Dustin Well-Known Member

    Started DTSC-660, Data and Database Management with SQL. This is a heavy course. There is around 20 hours of video content for this course, in addition to 5 quizzes and 4 assignments. Quizzes can be repeated, assignments are one-shot. 56% of the grade is based on the quizzes and 44% based on the assignments.
     
    SteveFoerster likes this.
  19. Dustin

    Dustin Well-Known Member

    I'm a week into DTSC-660 (the database course.) The videos are, so far, readings from a writeup that we are provided along with additional material from the textbook. So for now I'm focusing on doing that reading myself and drilling flashcards for the definitions. I've completed the first module and quiz, out of six.

    Module 1: Introduction to Database Management Systems
    Module 2: Database Design Processes
    Module 3: Introduction to SQL
    Module 4: Intermediate SQL
    Module 5: Triggers, Functions, and Procedures in SQL
    Module 6: Introduction to git and github

    Module 2 includes the first assignment and another quiz. The first assignment involves designing an Entity-Relationship diagram, and putting it into a set of relational schemas.
     
    SteveFoerster likes this.
  20. Dustin

    Dustin Well-Known Member

    The course I'll be taking next term, DTSC-600 Information Visualization is taught by Prof Javier Leon who I just learned is doing his PhD in Data Science at Northcentral. (https://www.ncu.edu/programs-degrees/doctoral/doctor-philosophy-data-science#gref) I'm really curious to learn how his experience is and whether that translates into him taking on a bigger role in the Eastern program.

    He has 3 Masters degrees (Food Marketing, Business Intelligence and Analytics, and an MBA) all from St. Joseph's University in Philadelphia.
     
    SteveFoerster likes this.

Share This Page