Skip to content
AI360XpertBeta
Glossary
Definition

Bayesian Linear Regression

Treating regression coefficients as probability distributions updated from a prior into a posterior as data arrives, so predictions come with data-driven uncertainty bounds.

Think of It Like This

Like drawing a pencil line on a graph that starts as a broad blurry cloud and sharpens into a tighter band as you plot more data points.

With a normal prior and normal noise, the posterior over each coefficient is exactly normal too, and its variance provably shrinks as more data accumulates. Taking the posterior's peak value, called MAP estimation, with a normal prior produces the same point estimate as ridge regression — ridge's L2 penalty is a normal prior in disguise.

The real payoff over a point estimate is predictive uncertainty: a full distribution over plausible outcomes for a new input, not just one best-guess number.