How Fermi Problems can Boost your Data Science Career
An article on the importance of learning how to estimate
It was the year 2011, and I was in my second year at the faculty of Mechanical Engineering in Pisa (oh, yes: I know you know the “leaning tower”!). One of our professors wanted us to experience the so-called “Fermi problems” so that we can learn how to estimate.
In the beginning, it was very difficult: we live in a world where everyone wants 0 error; and, furthermore, I was not good at estimating things, because before estimating you have to be curious enough. Then, I started getting the right pleasure from this kind of exercise, and curiosity started arising inside me.
Having the capability to estimate is a good skill if you are (or want to be) an Engineer or a Scientist, but in my opinion it is especially important if you are (or want to be) a Data Scientist, and in this article I want to explain you why (and how to solve Fermi problems).
1. What a Fermi problem is and how to solve it
Let’s start with a definition (quoting and translating from the Italian Wikipedia):
in applied mathematics, a Fermi problem is a problem designed to teach how to estimate, the importance of clearly identifying the…
