Lesson 19 — Linear regression (closed form)

Fit a straight line with a formula

Linear regression fits a line y = slope × x + intercept through the data. With one feature there is an exact formula, no training loop needed. The slope = Σ(x − mx)(y − my) / Σ(x − mx)², where mx and my are the means. The intercept (where the line crosses x = 0) = my − slope × mx.

# points (1,3) (2,5) (3,7): slope 2, intercept 1 -> y = 2x + 1

Why it matters: linear regression is the simplest predictive model and still used daily for prices, forecasts and trends.