DaveWentzel.com All Things Data
Charter Member of the Microsoft Professional Program in Data Science
Want to be a "data scientist"? Don't know where to start? Try the Microsoft Professional Program in Data Science. I am a charter graduate. It's a curriculum of edX courses that really help you think differently about your software projects and business. I certainly won't be adding "Data Scientist" to my resume any time soon, but it is a good first step.
"Data Science" is THE hot new skill right now. But how do you get started especially if you don't have the math background? Try the Microsoft Professional Degree Program. I took the accelerated charter version of the curriculum and loved it.
This guy has a great blog post about the program. You start by taking about 10 edX courses. These aren't difficult. The courses are ordered EXACTLY how a data scientist solves problems using a deductive, iterative approach. You start with a course on visualization tools. Then move on to doing data analysis with Excel and SQL. Only then do you learn about probabilities and statistics. Then you learn core data science concepts and ML algorithms.
Only after all of that do you touch R and python. This is important...a good data scientist doesn't use R as the Golden Hammer for every problem. Learn the data and learn your industry first. Very refreshing.
The next course is how to develop intelligent solutions using tooling like Spark and webservices. This is critical. I have yet to meet a data scientist that can take her R model and deploy it to a webservice or embed it into a data stream. Her model becomes useless unless it can be integrated into at least Excel.
The most enjoyable course is on Azure ML. I like to call Azure ML the "GUI for R". Azure ML democratizes data science and machine learning. What a great tool! R can be a distraction to a OO programmer without a strong math background. AML abstracts away much of that and allows you to develop theories about your data quickly...yet allows you to use R, python, and your Stats PhD when you really need it.
Finally, there is a "capstone project" where you actually have to put your knowledge to use. You are given a business problem and you need to solve it end-to-end. You need to analyze the problem, learn the industry, determine what data is important, cleanse the data, build the predictive models, train and validate them, and build a predictive webservice. You are graded on how good your predictions are.
This is real world stuff folks. Not the typical 50 question Microsoft certification test you take at Vue after you grab some braindumps off the web. YOU WILL LEARN SOMETHING. Even if you've been doing data science for 20 years, you'll learn something. And what you learn will be practical and applicable to your daily life.
Certificates are not being issued for a few more months but here is my temporary cert:
I hold a lot of certifications. This one was fun and I actually learned something.
You have just read "Charter Member of the Microsoft Professional Program in Data Science" on davewentzel.com. If you found this useful please feel free to subscribe to the RSS feed.