dmm.BaseModel: Create a model using bulit-in conjugate models

In nsdumont/jDirichletMixtureModels: Dirichlet Mixture Models for Clustering using Julia backend

Description

Create an model object to be used in the dmm.cluster function, using the packages bulit-in conjugate models. Function dmm.model is an alternative method.

Usage

model <- dmm.BaseModel(typename, params)

Arguments

typename	A string. The name of the predefined conjugate prior you wish to use. Options listed under details. "MultivariateNormalModel" is the default.
params	A list of the hyperparameter values for the likelihood functions. Many model have default params values and thus can be made without passing any params. See documentation for what parameters a given model may take.
data	In lieu of explicit hyperparameter values, some models can infer good hyperparameter values from given data. This option is supported for MultivariateNormalModel and UnivariateNormalModel, as well as UnivariateNormalKnownSigma.

Details

Bulit-in models avaible are: "MultivariateNormalModel" (default), "UnivariateNormalModel", "UnivariateNormalKnownSigma", "UnivariateExponentialModel".

Value

A model object of type BaseModel which can be passed to dmm.cluster.

nsdumont/jDirichletMixtureModels documentation built on May 23, 2019, 2:51 p.m.