Creates a new GradientGaussianMixture instance from arrays of weights, means and covariances
Creates a new GradientGaussianMixture instance from arrays of weights, means and covariances
Array of weights
Array of mean vectors
Array of covariance matrices
Optimization algorithm. Defaults to GradintAscent
Creates a new GradientGaussianMixture instance from arrays of weights, means and covariances
Creates a new GradientGaussianMixture instance from arrays of weights, means and covariances
Array of weights
Array of mean vectors
Array of covariance matrices
Optimization algorithm. Defaults to GradintAscent
Fit a Gaussian Mixture Model (see https://en.wikipedia.org/wiki/Mixture_model#Gaussian_mixture_model).
Fit a Gaussian Mixture Model (see https://en.wikipedia.org/wiki/Mixture_model#Gaussian_mixture_model). The model is initialized using a K-means algorithm over a small sample and then fitting the resulting parameters to the data using this {GMMOptimization} object
Data to fit the model
Number of mixture components (clusters)
Optimization algorithm
number of samples processed per iteration
maximum number of gradient ascent steps allowed
log-likelihood change tolerance for stopping criteria
The K-Means model will be trained with k*pointsPerCl points
Number of iterations allowed for the K-means algorithm
Random seed
Fitted model
Creates a new GradientGaussianMixture instance initialized with the results of a K-means model fitted with a sample of the data
Creates a new GradientGaussianMixture instance initialized with the results of a K-means model fitted with a sample of the data
training data in the form of an RDD of Spark vectors
Number of components in the mixture
Optimizer object. Defaults to simple gradient ascent
The K-Means model will be trained with k*pointsPerCl points
Number of iterations allowed for the K-means model
Number of K-means models to try
random seed