Linear Regression, Line by Line
Fit, read and question a linear regression so you know exactly what its numbers mean
A taste of a lesson
My model says price = 50000 + 2000 * size. Does that mean adding a square metre adds 2000 to the price?
Close, with two careful changes. First, it is an association: homes that are one square metre larger tend to be priced about 2,000 higher, in your data. Building an extension is a different question, because size travels with location and age. Second, it only holds within the sizes you observed. If your homes range from 40 to 150 square metres, do not use it for a 400 square metre villa. Try saying what the 50,000 intercept means, and whether it is meaningful here.
Written by the teacher as an example. In your lesson the tutor answers your own questions, and like any AI it can be wrong.
What you will be able to do
- Explain what least squares minimises and why outliers matter
- Interpret coefficients in a full sentence with units
- Encode a categorical feature and read its coefficients
- Judge R squared and residual plots without being misled
- Recognise when extrapolation or correlated features make results unreliable
Lesson plan
- 1 A line through five points Find a sensible line by eye, then see what least squares changes. Start
- 2 Reading slope and intercept State what each fitted number means for a real dataset. Start
- 3 Several features at once Interpret multiple regression coefficients holding other features constant. Start
- 4 Categories and one hot encoding Include categorical features and read them against a reference group. Start
- 5 R squared and residual plots Judge fit with plots first and summary numbers second. Start
- 6 Where linear regression breaks Recognise outliers, correlated features and extrapolation risks. Start
Try asking
About this tutor
A beginner tutor for people who want to really understand the most widely used model in data work. You start with five data points and a ruler, find the least squares line by reasoning, then move to several features, categorical variables and the meaning of each coefficient. Lessons cover reading a fitted model, R squared and its traps, outliers, correlated features and the danger of extrapolating. Every lesson ends with a small dataset to interpret in your own words. No heavy algebra: the maths appears only where it helps you read the output correctly.
Reviews
4.3
4 ratingsSample
- Hamid S.Sample
I finally understand why my coefficients flipped sign when I added a correlated feature. Good, honest explanations.
- Grace N.Sample
Clear and unhurried. The one hot encoding lesson was the most useful for me. A little light on how to check residual plots in practice.
- Erin P.Sample
Solid basics, but I already knew most of it. Better for true beginners than for people who have run regressions before.
- Tomasz K.Sample
The 'say it with units' habit fixed how I present results at work. Moving one point and watching the line tilt made outliers make sense.
About the teacher
Classical machine learning models, worked through on paper before any code
9 tutors 379 lessons taught Sample
I teach the classical machine learning models: regression, trees, ensembles, nearest neighbours, support vector machines and naive Bayes. My background is in applied analytics, where I spent a lot of time explaining to colleagues why a model made a particular prediction, and that shaped how I teach. I like to start with a tiny dataset you can hold in your...
See Kavya's profile and tutorsMore like this
Other tutors on the same or nearby topics.