The Dirichlet Distribution: A Density on the Simplex
An interactive live film: the Dirichlet density as a 3D surface over the triangle of probability vectors, flat at alpha = 1, a dome above one, climbing to the corners below one; the green line stands on the mean. Set alpha to hit a target mean and variance.
Narration
A probability vector over three categories is a point of this triangle on the floor: its coordinates are nonnegative and sum to one. This is the simplex.
The Dirichlet distribution puts a density on the triangle: a constant made of Gamma functions, times each coordinate raised to its concentration minus one. The height of the surface is that density.
The rest of the narration is in the book's Companion.
More from Chapter 13
Interactive1:35Normalizing Flows: Chaining Invertible MapsChapter 13
Preview1:38Independence as an Outer ProductChapter 13
Preview1:40Conditional Gaussian: A Slice of the BellChapter 13
Preview1:22Random Fourier FeaturesChapter 13
Preview1:46Change of Variables: Same Probability, Different AreaChapter 13
Preview1:35The von Mises-Fisher Distribution: A Gaussian for DirectionsChapter 13
Preview1:14Gaussian Mixture Model: Multimodal TerrainChapter 13
Preview1:41Whitening: PCA and ZCAChapter 13
Preview1:36Beta and Bernoulli: Updating with Pseudo-CountsChapter 13
Preview1:43ROC Curve and AUC: the Mass Above the DiagonalChapter 13
Preview1:40Hoeffding's Inequality: How Fast a Sample Mean SettlesChapter 13
Preview1:25Temperature Scaling: Calibrating a SoftmaxChapter 13
Preview1:34Mahalanobis Distance: The Ellipsoids of a GaussianChapter 13
Preview2:09Random Directions Are Nearly OrthogonalChapter 13