Learn R Programming

condmixt (version 1.1)

Conditional Density Estimation with Neural Network Conditional Mixtures

Description

Neural network conditional mixtures are mixture models whose parameters are predicted by a neural network. The mixture model can thus change its parameters in response to changes in predictive covariates. Mixtures included are gaussian, log-normal and hybrid Pareto mixtures. The latter relies on the generalized Pareto distribution to account for the presence of large extreme events. The unconditional mixtures are also available.

Copy Link

Version

Install

install.packages('condmixt')

Monthly Downloads

31

Version

1.1

License

GPL-2

Maintainer

Julie Carreau

Last Published

May 11th, 2020

Functions in condmixt (1.1)

condmixt.init

Conditional mixture parameter initial values
condgaussmixt

The conditional Gaussian mixture distribution
condmixt.fit

Maximum likelihood estimation for conditional mixture parameters
gaussmixt

Mixture of Gaussians
condhparetomixt.cvtrain.tailpen

Cross-validation of the conditinal mixture with hybrid Pareto components with a tail penalty added to the negative log-likelihood for training.
condmixt.nll

Negative log-likelihood for conditional mixtures
condmixt-package

Conditional Density Estimation with Neural Network Conditional Mixtures
condhparetomixt

The conditional hybrid Pareto mixture distribution
hpareto.negloglike

Hybrid Pareto Maximum Likelihood Estimation
condmixt.foldtrain

Training of conditional mixtures and evaluation of the negative log-likelihood on validation data
hparetomixt

Mixture of hybrid Paretos
condmixt.fwd

Forward pass in neural network conditional mixtures
hpareto

The Hybrid Pareto Distribution
gpd.mme

Moment Estimator for the Generalized and the Hybrid Pareto Distribution
hpareto.alpha

Auxillary Parameters of the Hybrid Pareto Distribution
hparetomixt.negloglike

Maximum Likelihood Estimation for a Mixture of Hybrid Paretos
gaussmixt.init

Provides initial values for the parameters of a mixture of Gaussians based on a sample.
hparetomixt.disp

Display the Hybrid Pareto Mixture Parameters
hparetomixt.init

Provides initial values for the parameters of a mixture of hybrid Paretos based on a sample.
hparetomixt.negloglike.tailpen

Maximum Likelihood Estimation for a Mixture of Hybrid Paretos with Tail Penalty
hillest

Hill Estimator
kneigh.condquant

Conditional quantile estimation from nearest neighbors.
lambertw

Lambert W Function
condmixt.quant

Quantile computation for conditional mixtures.
condmixt.train

Training of conditional mixtures
lognormixt

Mixture of Log-Normals
softplus

Softplus Transform
condmixt.dirac.negloglike

Negative log-likelihood for conditional mixture with a discrete component at zero.