daltoolboxdp (version 1.2.737)

autoenc_variational_e: Variational Autoencoder - Encode

Description

Creates a deep learning variational autoencoder (VAE) to encode sequences of observations. Wraps a PyTorch implementation.

Usage

autoenc_variational_e(
  input_size,
  encoding_size,
  batch_size = 32,
  num_epochs = 1000,
  learning_rate = 0.001
)

Value

A autoenc_variational_e object.

Arguments

input_size

input size

encoding_size

encoding size

batch_size

size for batch learning

num_epochs

number of epochs for training

learning_rate

learning rate

References

Kingma, D. P., & Welling, M. (2014). Auto-Encoding Variational Bayes.

Examples

if (FALSE) {
# Requirements: Python with torch installed and reticulate configured.

# 1) Create sample data
X <- matrix(rnorm(1000), nrow = 50, ncol = 20)

# 2) Fit VAE encoder
ae <- autoenc_variational_e(input_size = 20, encoding_size = 5, num_epochs = 50)
ae <- daltoolbox::fit(ae, X)

# 3) Transform to latent encodings
#    Note: the underlying Python returns [mean | var] concatenated; depending on
#    the implementation, you may receive 2*encoding_size columns.
Z <- daltoolbox::transform(ae, X)
dim(Z)
}

# See:
# https://github.com/cefet-rj-dal/daltoolbox/blob/main/autoencoder/autoenc_variational_e.md