to_keras.ruta_loss_named: Obtain a Keras loss

In ruta: Implementation of Unsupervised Neural Architectures

Obtain a Keras loss

Description

Builds the Keras loss function corresponding to a name

Usage

## S3 method for class 'ruta_loss_contraction'
to_keras(x, learner, ...)

## S3 method for class 'ruta_loss_correntropy'
to_keras(x, ...)

## S3 method for class 'ruta_loss_variational'
to_keras(x, learner, ...)

## S3 method for class 'ruta_loss_named'
to_keras(x, ...)

Arguments

x	A "ruta_loss_named" object
learner	The learner object including the keras model which will use the loss function
...	Rest of parameters, ignored

Value

A function which returns the corresponding loss for given true and predicted values

References

• Contractive loss: Deriving Contractive Autoencoder and Implementing it in Keras

• Correntropy loss: Robust feature learning by stacked autoencoder with maximum correntropy criterion

• Variational loss:

  • Auto-Encoding Variational Bayes

  • Under the Hood of the Variational Autoencoder (in Prose and Code)

  • Keras example: Variational autoencoder

ruta documentation built on Jan. 9, 2023, 1:20 a.m.