Struct SelfSupervisedAutoencoder 

pub struct SelfSupervisedAutoencoder<T: FloatBounds + ScalarOperand> { /* private fields */ }

Autoencoder for Self-Supervised Representation Learning

Implements simple autoencoder architecture for unsupervised feature learning.

Implementations

impl<T: FloatBounds + ScalarOperand + Debug> SelfSupervisedAutoencoder<T>

pub fn new( input_dim: usize, hidden_dims: Vec<usize>, config: AutoencoderConfig<T>, ) -> Result<Self, SklearsError>

Create a new self-supervised autoencoder

pub fn encode(&mut self, input: &Array2<T>) -> Result<Array2<T>, SklearsError>

Encode input to latent representation

pub fn decode(&mut self, latent: &Array2<T>) -> Result<Array2<T>, SklearsError>

Decode latent representation to output

pub fn forward(&mut self, input: &Array2<T>) -> Result<Array2<T>, SklearsError>

Forward pass through autoencoder

pub fn reconstruction_loss( &self, input: &Array2<T>, output: &Array2<T>, ) -> Result<T, SklearsError>

Compute reconstruction loss

pub fn sparsity_penalty(&self, latent: &Array2<T>) -> Result<T, SklearsError>

Compute sparsity penalty

Trait Implementations

impl<T: Clone + FloatBounds + ScalarOperand> Clone for SelfSupervisedAutoencoder<T>

fn clone(&self) -> SelfSupervisedAutoencoder<T>

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl<T: Debug + FloatBounds + ScalarOperand> Debug for SelfSupervisedAutoencoder<T>

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.