NeuralNet

class viabel.NeuralNet(layers_shapes, nonlinearity=<function primitive.<locals>.f_wrapped>, last=<function primitive.<locals>.f_wrapped>, mc_samples=10000, seed=1)

Attributes:

dim

Dimension of the space the distribution is defined on

supports_entropy

Whether the approximation family supports closed-form entropy computation.

supports_kl

Whether the approximation family supports closed-form KL divergence computation.

var_param_dim

Dimension of the variational parameter

Methods

entropy(var_param)	Compute entropy of variational distribution.
init_param()	A variational parameter to use for initialization.
kl(var_param0, var_param1)	Compute the Kullback-Leibler (KL) divergence.
log_density(var_param, x)	The log density of the variational distribution.
mean_and_cov(var_param)	The mean and covariance of the variational distribution.
pth_moment(var_param, p)	The absolute pth moment of the variational distribution.
sample(var_param, n_samples)	Generate samples from the variational distribution
supports_pth_moment(p)	Whether analytically computing the pth moment is supported

forward

__init__(layers_shapes, nonlinearity=<function primitive.<locals>.f_wrapped>, last=<function primitive.<locals>.f_wrapped>, mc_samples=10000, seed=1)

Parameters:

layers_shapeslist of int

The hidden layers dimensions.

nonlinearityfunction

Non linear function to apply after each layer except the last layer.

lastfunction

Non linear function to apply after the last layer.

mc_samplesint

Number of samples to draw internally for computing mean and cov.

seedint

Internal seed representation.

log_density(var_param, x)

The log density of the variational distribution.

Parameters:

var_paramnumpy.ndarray, shape (var_param_dim,)

The variational parameter.

xnumpy.ndarray, shape (dim,)

Value at which to evaluate the density.

mean_and_cov(var_param)

The mean and covariance of the variational distribution.

Parameters:

var_paramnumpy.ndarray, shape (var_param_dim,)

The variational parameter.

sample(var_param, n_samples)

Generate samples from the variational distribution

Parameters:

var_paramnumpy.ndarray, shape (var_param_dim,)

The variational parameter.

n_samplesint

The number of samples to generate.

Returns:

samplesnumpy.ndarray, shape (n_samples, var_param_dim)

supports_pth_moment(p)

Whether analytically computing the pth moment is supported