Beta and Bernoulli: Updating with Pseudo-Counts
An interactive live film: one Beta posterior per coin toss, standing one behind another; heads add to alpha, tails to beta. Find the prior strength at which twenty tosses barely move the mean.
Narration
A coin lands heads with an unknown probability p. Our belief about p is a beta distribution: a density on the interval from zero to one, here the front curve.
Its two parameters, alpha and beta, act like counts of heads and tails we have already seen. Their mean is alpha over alpha plus beta, and the more pseudo counts, the narrower the curve.
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
Preview1:42The Dirichlet Distribution: A Density on the SimplexChapter 13
Preview1:14Gaussian Mixture Model: Multimodal TerrainChapter 13
Preview1:41Whitening: PCA and ZCAChapter 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