Understand Basic Linear Regression Concepts To Get Started With Machine Learning

Why it’s called “regression,” how to find the regression line (intuition), and how to measure how good it is

Photo by Chris Lawton on Unsplash.

It’s time to take two steps back in my machine learning journey and dive into the math, algorithms, and intuition behind the scenes. It’s been fun to learn frameworks such as PyTorch, TensorFlow, and the Python language itself. But I realize I’m always getting stuck and having to circle back to understand things better. I want to close this gap.

I’ll begin with writing about linear regression. I won’t cover the various calculations and formulas too deeply. Instead, I want to focus on building intuition and teaching the basics.

Linear regression is a modelling approach and algorithm used in machine learning to assess the correlation between one dependent variable and one or more independent ones. Here are a few examples:

• Years of work experience and salary — If you want to predict a person’s salary from their experience, then “Salary” is the dependent variable y (the vertical/y-axis in a graph) and “Experience” is the independent x variable (the horizontal/x-axis in a graph). It’s possible to use other independent variables (also known as explanatory…