When you think of a data scientist, you might think of someone who plays Minecraft, attends Star Trek conventions and wears taped glasses. Maybe he looks like the guy to your left. Well, our lead data scientist challenges that stereotype. Kind of.
Meet Michael Faron. He’s a badass.
While Michael may dabble in the occasional video game (BioShock is his favorite right now) own a Captain Kirk t-shirt and, okay, he sports some black-rimmed glasses, he’s truly the coolest data scientist you’ll ever meet.
In addition to being a ninja with numbers, he’s also a screenwriter, mountain biker and photographer. For your viewing pleasure, I’m going to thread his photography throughout this post--starting with the example to the left. And don’t worry. There’s more where that came from. You can check out his portfolio here.
Because of his rare left and right brain harmonization—maybe we could call him ambi-cerebrus—he’s not only good at data science, he’s good at explaining it.
So, here’s Michael simplifying the complex arts of data science and predictive modeling.
In three sentences or fewer, describe predictive modeling.
I like to think of predictive modeling as very similar to when you or I make an educated guess. We may have good information to help us make a decision; even though we know there is a chance our guess might be wrong. Predictive modeling attempts the same thing but leverages computers and complex algorithms to search through and analyze data in the hopes of discovering useful patterns hidden inside. With the right data and the right techniques, predictive modeling can generate effective and useful decisions.
Can you give us an analogy for predictive analytics?
It’s the fourth leg on the barstool of decision-making. So, traditionally you make decisions based on three dimensions: 1.) Education 2.) Experience 3.) Instinct. Envision those decision-making dimensions as three legs on a barstool. The barstool will work—but it’s not as sturdy as a barstool with fourth legs. Analytics is that fourth leg or dimension that adds a new level of assurance to your decision-making process.
What are different ways we interact with predictive modeling on a daily basis?
Here are five ways everyone will be familiar with:
1. Google. Every time we enter something in Google’s search bar, the results we get back come from complex predictive analytics processes including language and text analytics.
2. Advertisements. The advertisements we see on the web or receive in the mail often use predictive analytics to make guesses about our interests by evaluating past behavior or associations.
3. Loans. Banks use credit scores (like FICO) to make an educated about the risk of lending money.
4. Insurance. Our car and medical insurance rates and premiums are designed around statistical models and predictive analytics.
5. Netflix and Amazon. These companies use predictive analytics to make recommendations about movies or products their customers might be interested in by analyzing past data or evaluating similarities customers have to each other.
Why should we all love predictive modeling?
One of the greatest benefits of advanced analytics is that it provides opportunities to learn and understand in ways not previously possible. As the sheer volume and complexity of data continues to grow, it can outpace our ability to analyze and understand it. But as computing systems and analytics techniques continue to evolve, we gain the tools and processes that allow us to extend our reach beyond where we are today in order to explore, discover and solve problems in new and different ways. It’s really exciting.
*photo credit: uncyclopedia commons