Struct MixUp 

pub struct MixUp<A: Float> { /* private fields */ }

MixUp augmentation

Implements MixUp data augmentation, which linearly interpolates between pairs of examples and their labels, helping improve model robustness.

Example

use scirs2_core::ndarray::array;
use optirs_core::regularizers::MixUp;

let mixup = MixUp::new(0.2).unwrap();

// Apply MixUp to batch of inputs and labels
let inputs = array![[1.0, 2.0], [3.0, 4.0]];
let labels = array![[1.0, 0.0], [0.0, 1.0]];

let (mixed_inputs, mixed_labels) = mixup.apply_batch(&inputs, &labels, 42).unwrap();

Implementations

impl<A: Float + Debug + ScalarOperand + FromPrimitive + Send + Sync> MixUp<A>

pub fn new(alpha: A) -> Result<Self>

Create a new MixUp augmentation

Arguments

• alpha - Parameter for Beta distribution; larger values increase mixing

Errors

Returns an error if alpha is not positive

pub fn apply_batch( &self, inputs: &Array2<A>, labels: &Array2<A>, seed: u64, ) -> Result<(Array2<A>, Array2<A>)>

Apply MixUp to a batch of examples

Arguments

• inputs - Batch of input examples
• labels - Batch of one-hot encoded labels
• seed - Random seed

Returns

Tuple of (mixed inputs..mixed labels)

Trait Implementations

impl<A: Clone + Float> Clone for MixUp<A>

fn clone(&self) -> MixUp<A>

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl<A: Debug + Float> Debug for MixUp<A>

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

Formats the value using the given formatter.

impl<A: Float + Debug + ScalarOperand + FromPrimitive, D: Dimension + Send + Sync> Regularizer<A, D> for MixUp<A>

fn apply(&self, _params: &Array<A, D>, gradients: &mut Array<A, D>) -> Result<A>

Apply regularization to parameters and gradients

fn penalty(&self, params: &Array<A, D>) -> Result<A>

Compute the regularization penalty value