liveBook · Manning

Gamma distribution

The gamma distribution is a continuous probability distribution over positive values, parameterized by shape and scale or rate.

How 3 Manning books explain it

Ensemble Methods for Machine Learning

A gamma distribution is a continuous probability distribution on positive values commonly used to model waiting times or right-skewed magnitudes.

How it is used
Models continuous, positive, right-skewed regression targets such as insurance claim amounts.
How deep it goes
Explains support, tail behavior, surrogate likelihood, GLM formulation, and alternatives, but does not derive the gamma density.
Examples
Uses insurance claims, long-tailed positive targets, and bike-sharing counts to contrast gamma with Tweedie modeling.
What you can do after
Apply it via gamma regression, LightGBM, or XGBoost when targets are positive, continuous, and right-skewed.

Read how Ensemble Methods for Machine Learning explains gamma distribution →

Math and Architectures of Deep Learning

A probability distribution over positive continuous values, here mentioned as the univariate distribution previously used in Bayesian parameter estimation.

How it is used
Uses gamma distributions as conjugate priors and posteriors for Gaussian precision when variance is unknown.
How deep it goes
Defines the density and derives posterior parameters and the MAP mode, while deferring the gamma function’s full introduction to an appendix.
Examples
Works through Gaussian precision estimation with known mean and supplies PyTorch posterior-update code.
What you can do after
Apply it via PyTorch to compute a gamma posterior and MAP estimate for Gaussian precision from observed data.

Read how Math and Architectures of Deep Learning explains gamma distribution →

Grokking Bayes

The gamma distribution is a continuous probability distribution over positive values, parameterized by a shape and a rate or scale.

How it is used
Models an unknown event rate as a Bayesian prior before historical occurrence data refine the estimate.
How deep it goes
Defines the distribution and describes prior updating, but does not develop its parameterization or calculations.
Examples
Uses call-center arrivals, soccer goals, and insurance claims to illustrate unknown event rates.
What you can do after
Evaluate unknown event rates and refine beliefs using historical occurrence counts.

Read how Grokking Bayes explains gamma distribution →

Also covered in

Related concepts

Questions about gamma distribution

What is gamma distribution?

The gamma distribution is a continuous probability distribution over positive values, parameterized by shape and scale or rate.

Which Manning books explain gamma distribution?

Ensemble Methods for Machine Learning, Math and Architectures of Deep Learning and Grokking Bayes.

What is the difference between gamma distribution and Wishart distribution?

The gamma distribution is a continuous probability distribution over positive values, parameterized by shape and scale or rate. The Wishart distribution is a probability distribution over symmetric positive-semidefinite matrices, commonly used to model covariance-matrix estimates.

These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.