How the Formula 1 Can Help You to Understand Under and Over Fitting
When to Identify It and What You Can Do About It
Introduction
I sometimes feel like I have fluked the system. Countless exams have tested my ability to identify and explain regression modelling techniques, and if I had to look at my answers now, I would be amazed at my knowledge. However, regurgitating your notes is a life-hack that does not stand the test of time. Sometimes concepts just Do-Not-Stick. And there is something you can do about it.
One of these concepts was (past tense) the bias-variance trade-off — also known as — over and underfitting. When did I realize this? When I was asked to explain it to someone who didn’t have an analytical background. I have my MSc in Econometrics but found it difficult to explain a basic principle to my neighbour. So, I took on this challenge and I am here to share it with you.
Overfitting and Underfitting — An Analogy:
You need to prepare for your driving test to get your license. How can one better prepare than being trained by the best drivers in the world? So, you decide to analyze the Formula 1(F1) drivers every Sunday — logical right?
