Linear Regression
Linear regression draws a straight line through dots so it can guess a number.
What it does
Linear regression draws a straight line through dots so it can guess a number.
Regression: Predicting a number • Fit: How close the line is
How it works
- Put the example dots on a chart.
- Draw a line that stays close to the dots.
- Use the line to guess a new number.
Uses and applications
- A weather app can guess tomorrow’s temperature from past temperatures.
- A home website can estimate a house price from size and location.
- A growth chart can estimate a child’s height at a future age.
In short
It predicts numbers by drawing a best-fit line, like a growth chart.
Related methods
- Logistic Regression: Uses a similar weighted sum, but turns it into a chance for a yes-or-no answer.
- ARIMA: Also a linear model, applied to values ordered in time.
Back to Prediction and classification · All topics
The interactive demonstration and the Python walkthrough for this topic need JavaScript.