Struct FeedForwardNet 

pub struct FeedForwardNet {
    pub layers: Vec<DenseLayer>,
}

Feed-forward neural network with multiple dense layers.

Fields

layers: Vec<DenseLayer>

Implementations

impl FeedForwardNet

pub fn new(layers: Vec<DenseLayer>) -> Self

Create a new network from a list of layers.

pub fn forward(&self, input: &DVector<f64>) -> f64

Forward pass: compute scalar output from input vector.

pub fn forward_with_intermediates( &self, input: &DVector<f64>, ) -> Vec<DVector<f64>>

Forward pass returning all intermediate activations (for backprop).

pub fn backward(&self, input: &DVector<f64>) -> DVector<f64>

Backward pass: compute gradient of output w.r.t. input.

pub fn input_dim(&self) -> usize

Input dimension expected by the network.

Trait Implementations

impl Clone for FeedForwardNet

fn clone(&self) -> FeedForwardNet

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for FeedForwardNet

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

Formats the value using the given formatter.