A Friendly Introduction to Data Science

This article was originally published on https://technocampos.com

There’s a new trend in the technology industry called Data Science. It’s on the news, people have heard about it, but is hard to grasp what it actually means.

Even though the concept is “new”, it is not so complicated as it looks.

Data Science – as used now – is nothing more than the good old scientific method (science) applied to the ever increasing data generated by the techno beast. All in all, this “new field” is just an extrapolation of the old model to our new context.

Without computers Data Science is just good old research

The main difference between Galileu Galilei and a 21st century Data Scientist is that Galileu didn’t have massive amounts of data to back his hypothesis up, nor computers to analyze it all. Just remember that Newton used the fall of an apple to formulate gravitation’s mechanics!

Thales of Miletus, the father of western science, simply decided to explain the world based on natural terms instead of Greek gods. And just like that, based on questions, observations and hypothesis, the stepping stones of our high tech scientific society were laid. Fast forward 2,500 years and the essence of the scientific method is still the same. It is so simple that even children apply it on a daily basis.

What’s not so simple is the complexity of the problems and resources that we have now. Instead of one head and an apple, Google has a gazillion bytes to be analyzed by tens of thousands of engineers. The scope of the these problems is so big that it is impossible for humans to figure out by themselves. And to be honest, not even computers can interpret it all. When creating models, data scientists often reduce and approximate the original datasets they use to make the models computationally viable and humanly interpretable. This is not really different than a high school physics problem encouraging to ignore air resistance.

Nature’s interactions are complex, but sometimes it is possible to extract simple models off of it.

F = ma

Sometimes is not:

Hydraulic Jump Equation