My first job in the Navy was pretty awesome: I was a support guy attached to a Navy Special Operations team based in Virginia Beach, VA.
Now, before I say anything else, let me get some disclaimers out of the way: I was not a SEAL. I never went to BUD/S, and I was never a special operator in any sense of the word. I was a support guy. I knew my role, I loved it, and I did it well.
Many of the things that I learned and observed during those three years still stick with me today. One of them is the profundity of a statement that ‘those guys’ sometimes used during training evolutions: “Slow Is Smooth, and Smooth is Fast.” It may sound a little strange at first — even paradoxical, perhaps — but the idea is that when you’re learning a new skill, you need to avoid the temptation to rush.
This temptation is completely natural, by the way. After all, you’re trying to accomplish something of course, you’d rather do it sooner than later. Perhaps others can already do it, and you hope to catch up. Or, perhaps you’re just strongly motivated by the desire to master some element of your craft. The problem is, if you just fly through in order to rush to some imagined finish line, the important concepts aren’t likely to stick. On the contrary, going slowly at first — actually taking the time to look around, to think about what you’re doing, and to get it right — is what puts you on the true path towards mastery.
As a Data Science instructor, I think about this quote all the time, and I share it with my students.
Data Science is, to put it mildly, a hot field. Even within Data Science, there are certain modeling processes and techniques that carry a special sort of cachet — things like random forests, boosted trees, ridge regression, neural networks, etc. It’s natural for people to want to rush into these things, but it’s inadvisable to do so without first actually knowing what they mean, and without taking the time to understand the building blocks.
The same thing is true for the R and Python code that is often used to generate models using these algorithms.
Way too often, students want to race to the finish line — they want to run some lines of code through their console to be able to drop a hot buzzword onto their resume or into a conversation, only to become frustrated when all they see back is a syntax error.
If you’re starting out in Data Science, just remember that you’re running a marathon — not a sprint. Remember to pause for a deep breath. Remember that there are innumerable resources online that can help to deepen and to broaden your understanding of related topics.
Someone who calls himself a data analyst, or — gasp — a data scientist — but who can’t explain ‘how’ or ‘why’ a model generated the results that it did is really just a typist. They’ve typed letters and numbers into a computer keyboard accurately, and by doing so, caused some result to occur. Certainly, no offense is meant here to typists — but that’s not exactly a specialized skill set in this day and age.
To raise your game to the next level, just remember the expression at the top of this post. If you need to, say it over and over again, like a mantra. Then, put it into practice. If you’re not sure about what some function in R or Python does? Pull up the help menu. Read it. Slowly. If there are terms you don’t know? Look them up. Learn them. If you’re not sure how some value was calculated? Find a video tutorial. Watch it. Then watch it again.
If you can’t be bothered by the details, and you’re always rushing ahead to chase the pot of gold at the fastest speed possible, you’ll never actually accomplish your goals in data science.
GoodWaveData 