University of Pittsburgh Master of Data Science

I have been interested in getting a Masters degree in Data Science for a while now. Most of my delay in moving forward has been due to cost of the advanced degree versus an actual financial return for my particular situation. Up until now, most of these programs cost at least $50k and there is a long drawn out process for applying and I just didn't see a good return on my investment. This program changes all of that.

The first thing that attracted me to the program was the cost. It's only $15k for the entire program and that is very affordable for an advanced degree that is from a top R1 research institution. In addition, the program doesn't require going through the normally time-consuming process of applying, sending in transcripts, letters of recommendation, etc. It's very simple. They automatically verify your Bachelor's Degree on their end (unless you are an International student), and you simply take your first course. The magic of this process is that this first course acts as your application and acceptance all in one package. As long as you pass the course with an 80 or higher, you are then officially accepted into the program. This makes things so much easier and convenient for someone like myself.

So, why did I choose to move forward with this program?

• Affordable ($15k for entire program)
• Online - no interest in going into an actual physical classroom environment
• In depth learning of the material - the depth and coverage is fantastic
• Great educational institution - University of Pittsburgh is a great R1 Research University - It's not Ivy League but for the cost it's a great value

Now let's talk a little about the first course. It's called Introduction to Data-Centric Computing and I'm currently on Week 6 of the program. The first week was a lot of items packed into a short time frame. We covered the following:

• Basic Python and Jupyter Notebooks
• Setting up your programing environment, primarily focusing on Python and Jupyter, using environments, etc.
• Data types - Strings, Booleans, Lists, etc and indexing and slicing
• For loops

The second week was a ton of information about using basic Python to do calculations based around random number generation and various types of averages. We covered the following:

• Controlling randomness with the random module and using seeding
• In depth coverage of iteration - more For loops, Nested For loops, Iterating variables, etc.
• Repeating randomness and lots of work with List Comprehensions
• Working with averages - generating data, summing and counting, defining a function to compute the average, comparing the average of multiple samples
• A ton of work on can we trust the average? - simulations, sample size, variability, bell curve, effect of sample size, Central Limit Theorem and variance, Standard Deviation, Standard Error of the Mean, Understanding Standard Error, Confidence Intervals

Going into week 3 was pretty exciting for me because a lot of week one and week two was review for me, either from previous college coursework or just daily stuff. Week 3 we start to move beyond basic Python manipulation and into the world of Python packages and modules. These greatly enhance the experience when working with data and week 3 was dedicated to NumPy and how it's much easier to do most things when we use NumPy versus standard Python. We covered the following:

• Dictionaries versus lists
• Counting - unique values and repeated items, Dictionary comprehensions
• Side effects - list and dictionary specific
• Errors and debugging - including an intro to the Python debugger
• NumPy 1D Arrays - Arrays vs lists, slicing, combining, sorting, and summarizing
• NumPy 2D Arrays - Manipulating, slicing, indexing, subsetting and summarizing
• NumPy Random Numbers
• NumPy Standard Error on the Mean

Week 4 was even more interesting for me as it introduced Pandas. Using Pandas and NumPy makes all of the work required when doing the calculations with basic Python so much quicker and easier. We covered the following:

• Pandas Series - basic and practical application, indexing and slicing, concatenation
• Pandas DataFrames - conversions, attributes, methods, statistics and sorting
• DataFrames indexing - row selection, column renaming, index manipulation
• Working with columns and rows and analysis
• DataFrame filtering and operations
• Visualizing randomness with Pandas - box plots and sample size analysis

Week 5 was a deeper dive into Pandas with a lot of in-depth coverage of the following:

• Filtering - using strings, creating DataFrames, row filtering by value, row filtering by pattern matching
• Reading data into Pandas - excel, CSV, web and existing modules
• Combining DataFrames - concatenation, joins, join types, chaining multiple joins
• Exploring data - summarizing, SEM, missing values and axis considerations, Exploratory Data Analysis, group aggregation, split-apply-combine

I'm currently on Week 6. As we go more into areas that I haven't worked a lot with before, I'm pretty excited to cover all of the upcoming material. This is what I've covered so far in Week 6:

• Intro to MatPlotLib - creating figures and subplots, plotting a sine wave, multiple subplots, customizing subplots

I've really only just begun week 6 and there is a lot of info to cover this week, all based around MatPlotLib, counting, variables, and relationships. It looks very interesting and getting to start working with visualization tools is a nice change of pace. I've used Tableau a lot and MatPlotLib seems a lot more basic visually, but I'm finding that it's very powerful and looking forward to digging deeper.

In closing, I'm on week 6 of a 16 week program in the very first course of the Masters degree. So I'm literally just getting started. That being said, the amount of material that has been covered over the first 6 weeks has been amazing. The lectures have been extremely in-depth, and the cohort that I'm a part of has been interesting to work with as well. We use Slack and there is a section designed specifically by course. I will keep writing about the program until I graduate. It is a good opportunity for me to document what I'm learning and to give anyone who i s interested more insight into the program and whether or not it's right for them.