I can't decide if Coursera is the best or worst thing ever. I love the access to great classes. But, I could easily spend all my time learning instead of doing.
Earlier in the year, I completed the Data Analysis class. Last week, I started the Machine Learning class as a prelude to taking the full CS229 course in September (the last class I need to complete the Mining Massive Data Sets Graduate Certificate Program). Today, I started the High Performance Scientific Computing class. I'm also casually following Scientific Computing by just watching the videos. Needless to say, Coursera is consuming all of my spare time.
As much as I value learning, I believe that true understanding comes from practice. I need to find a balance between the two. This isn't a new challenge for me but at least I'm thrashing within the domain of data science instead of across different domains. That's progress of a sort.
My plan (read hope) is to apply Cal Newport's stretch churn concept by doing a series of small-but-challenging projects that stretch my skills along a particular dimension. I'm also hoping to apply two of Newport's other concepts, deep work and the textbook method.
It's a good plan. Now just the small matter of resisting the Coursera temptation.