When I was looking for Data Science certificate programs back in 2013, there were only a few available and most had only graduated one or two cohorts. Even worse, I could not find a single review for any of them. So, this is my review of the University of Washington Data Science certificate.

Background:
I ended up choosing the University of Washington program for a few reasons: it was part-time with only one 3-hour class per week, it was on-line, the application process was not very difficult, and I was able to get my employer to pay for it. I’ve got a wife, house, work full-time, and have all the normal bills and commitments that come with that. For the limited options available to me, this was an ideal fit.

I came into the program with a limited programming background and a very basic understanding of data science. Despite that, I never once felt like the material was that far above my head. For one, we used the programming language R, which was new to pretty much everyone in the class. In addition, the data science material we were learning was new to everyone. To summarize, I would say that even with a light technical background, you can still reap a lot of benefits from this program.

The Classes:
Intro to Data Science – This was the primer class where we learned the basic and intermediate capabilities of R – loading data, data frames, data manipulation, graphing, basic modeling, etc. Basic data science topics such as how to prep and clean data for training/testing, supervised learning, unsupervised learning, data models, etc.

Methods for Data Analysis – This class focused primarily on the statistics behind data science: inferential statistics, data distribution types, testing and experimental design, Bayesian vs. Classical statistics, linear/logistic regression, decision trees, SVM, visualization, etc.

Deriving Knowledge from Data at Scale – This final class was more of a review of the material from the first two classes. We also learned how to apply the material we had learned to real-world scenarios. The Capstone project for this class was to compete in a Kaggle competition and then present our findings.

General Thoughts:
Three hours of class per week plus homework was a pretty solid load while working full-time. It was a good commitment without feeling like too much and still left extra time to delve deeper into interesting subjects. I especially enjoyed the first two classes, there were plenty of useful assignments and I learned quite a bit. We used several different tools throughout the program such as R, Octave, Weka, and SQL. Assignments generally started easy towards the beginning of a class and got progressively more difficult. One thing that we didn’t really cover was Hadoop and other tools for processing really large datasets. This became an issue during our Kaggle competition since we were working with 10 GB + datasets. But with any new program there will be growing pains, so hopefully they’ve figured out how to incorporate this in the curriculum by now.

Conclusion:
Will this program leave you with all the skills and confidence to be gainfully employed as a Data Scientist? The short answer is NO. The long answer is that it will provide you with enough of a foundation and a lot of the tools to start doing some damage. It will be up to you to take that foundation and keep going from there.

Would I recommend this program? If you live in the Seattle area, are committed to working a full-time job, and have very little experience with data science, this class is for you. I’m not as keen to recommend this program to distance students because you won’t have direct access to the entire student and faculty network, which is very important in a collaborative field like Data Science. If you can, find a program that’s local. On top of all the learning you’ll do, you’ll develop a great network of like-minded people, which from my experience, will be key to your success.