Why Data ScienceSun 21 February 2016
Say you're the type of person that curious about how the world works.
In most cases you would need to spend years studying the subject, learning every nuance and detail. If you wanted to learn about Chess you might turn to Gary Kasparov or if you were interested in Go Champion Lee Sedol would be a good choice. Say you wanted to understand what made a drug safe for humans you might need to ask entire team of experts employed company like Merck with over a 100 years in the business.
No doubt all of these people and organizations would come to mind first. After all as folks they have demonstrated their ability to beat any other unaided human.
But what if a second set of people can outperform these individuals? And what if those people could do it in as little as a couple of months, moving from one problem to the next?
As you might have guessed the second set of people now exists. They are adept at using modern tools and mathematics to to rapidly understand the key drivers of complex systems, then create models and predictions that outperform any other human or system for the task at hand.
As it stands humans using computers are better at Chess, Go, and predicting which drugs are safe, better than the traditional experts in field. And on top of that these people are given enormous flexibility to learn from subject matter experts, answer difficult questions, and learn an enormous amounts about pretty much any topic of their choosing. And they can do this over and over for a wide range of subjects.
So there exists a job where you get to work with smart people, learn, and employers seem to give you a lot of flexibility to do it?
Well it turns out quite a bit. I won't get into the details here but typically you have to know a lot of math, be really good at statistics, know a programming language or two, know a lot (seriously a lot) about how data is captured, stored, transferred, and interpreted, and be an excellent communicator.
And then after all trying to learn all that stuff you have to convince one or more people that you're worth paying to do some or all of what's mentioned above. This is the data science interview.