If you’ve been following along, you’ve read that my data science learning plan started out with a combination of books, MOOCs, and Kaggle. After about 1.5 years of dabbling with these and a few small projects at work, I came to a crossroads in my learning plan. There were three options:
- Put this data science thing to the side and focus on my software engineering career.
- Continue with learning that was semi-ad hoc.
- Enroll in a master’s program or graduate certificate program.
Option 1) was quickly eliminated. The reasons that I started learning about data science still applied. The work that I had done only increased my appetite to learn about this field. It was time to double down, not bail out.
Option 2) was a consideration. It was the path of least resistance after all. But the next steps in the learning path weren’t obvious. Since data science was a relatively small part of my job, learning that was driven by on the job needs wasn’t going to be sufficient. Figuring out which external projects or classes to pursue and in what order was overwhelming. There have been attempts to build a curriculum out of online materials, but I wasn’t satisfied with what I found nor with what I sketched out.
Ultimately, making the investment in a master’s program was the option that won out. There were several reasons for this:
- I wanted a curriculum that provided a solid grounding in statistics and other areas that are the foundation of data science.
- I felt that the degree would give me some flexibility in my career choices down the road.
- I like new challenges… and going back to school after many years would be that.
- I worked for an employer that would pay for me to go to school (arguably the most underutilized benefit at many companies).
- I was planning a 12 week sabbatical that would enable me to jump into a program quickly.
- I had the time to invest and the support of my family to prioritize this activity.
The next step was to find a program that aligned with my needs. First and foremost, it needed to be online and something I could work on part-time. There were some options at the time, but they have significantly increased in the last three years.
There were a couple of graduate certificate programs that I looked into, but ultimately didn’t feel like they gave me the technical grounding that I wanted.
There were a few online programs that I looked into, but the University of Wisconsin’s program was the only one I seriously pursued. It had the right mix of targeted curriculum, affordability, and name brand recognition. With 9.5 of 12 classes under my belt, I’ve been very happy with the program. I have no regrets about deciding on the master’s route nor on the school choice.
I expect to write several posts on the topic of coursework in graduate school and Wisconsin’s program – key things that I’ve picked up as we well as gaps and challenges. I have some personal experience to share, but there are plenty of other perspectives available. For example, I came across a blog post this week by Rafael Irizarry that reflects on the different specializations within data science and academia’s ability to meet those needs (https://simplystatistics.org/2018/11/01/the-role-of-academia-in-data-science-education/). The bottom line is that the field is young and the academic programs are even younger, so there will be some maturing along the way.
Photo details: Oct 23, 2018, Full Moon Rising over Lida, Canon EOS Digital Rebel, f/4, 1/4 second, ISO 1600