Choosing the Best Regression Model: A Rollercoaster Ride through ML

Kevin Akbari
5 min readOct 21, 2024

Ever wondered how to choose the best regression model? Like that time you spent hours choosing the best pair of shoes, but when it comes to machine learning, you can’t exactly “try before you buy”? Fear not! We’re here to take a humorous yet educational plunge into the math, logic, and sheer absurdity behind choosing the best regression model. Buckle up, my friend, because we’re about to board the Regression Rollercoaster. 🚀

The Core of Regression

At its heart, regression is all about predicting numbers — could be anything from house prices to how many cats you’ll adopt in the next decade (five, by the way). But how do you choose the best algorithm for the task? You could go with your gut… but that might land you with a model about as reliable as predicting the weather using tea leaves.

Instead, let’s look at the options and the criteria for choosing a mathematically sound model. I promise I’ll keep the math relatively painless and throw in a few jokes along the way!

Meet the Contestants 🎭

1. Linear Regression: The Old-Timer

Ah, Linear Regression — simple, humble, and as old as my grandma’s toaster. It’s the foundation of regression and assumes a…

--

--

Kevin Akbari

I enjoy exploring data science and delving into cutting-edge models currently utilized in various industries. https://www.linkedin.com/in/kevinakbari/