Dirichletprocessbased
The Dirichlet process is a stochastic process that describes a distribution over probability distributions. It is a fundamental building block in Bayesian nonparametrics and is often used in areas like topic modeling, clustering, and density estimation. A Dirichlet process, denoted DP($\alpha$, $G_0$), is characterized by two parameters: a base distribution $G_0$ and a concentration parameter $\alpha$. The concentration parameter controls how likely it is to draw new distributions that are "close" to the base distribution. A higher $\alpha$ leads to distributions that are more spread out and tend to have more distinct components.
When a distribution $F$ is drawn from a Dirichlet process, $F \sim DP(\alpha, G_0)$, any sample $x$
The Dirichlet process is often used in conjunction with the stick-breaking construction or the Chinese restaurant