the Dirichlet distribution does: distribution over probability simplices

How do we predict the likelihood of various outcomes in uncertain situations?

Imagine you're a chef trying to predict how popular different dishes will be at your restaurant.

Think of the Dirichlet distribution as a recipe that helps you mix your predictions for each dish's popularity, considering how much you know about them so far.

Example

If you've tried 3 dishes and know 2 are popular, the Dirichlet distribution helps you predict the third dish's popularity by considering the popularity of the first two.

Remember this

The Dirichlet distribution helps you blend your predictions for different outcomes based on past experiences, ensuring a balanced and informed forecast.

Text adapted from Wikipedia, licensed under CC BY-SA 4.0.

Related concepts

Markov chain Monte Carlo

MCMC samples from complex posterior distributions

Entropy H = -Σ p(x) log₂ p(x) measures average surprise in bits

How do we measure uncertainty in everyday decisions?

Metropolis–Hastings algorithm

Metropolis-Hastings algorithm samples from difficult distributions

Bayesian inference

Ever wondered how doctors update diagnoses as new symptoms arise?

log-probabilities are used instead of probabilities: avoids numerical underflow

Why can't we just add up tiny chances over time?

cross-entropy equals negative log-likelihood for classification

Why does knowing the wrong probability help us measure information loss?

Swipe through 100 ML concepts daily

Open Pocket Polymath