Struct RandomizedSmoothingClassifier 

pub struct RandomizedSmoothingClassifier {
    pub sigma: f64,
    pub n_samples: usize,
    pub confidence: f64,
}

Fields

sigma: f64

n_samples: usize

confidence: f64

Implementations

impl RandomizedSmoothingClassifier

pub fn new(sigma: f64, n_samples: usize, confidence: f64) -> Self

pub fn smooth_predict<F: Fn(&[f64]) -> usize>( &self, x: &[f64], base_classifier: &F, ) -> usize

Smooth prediction: return class with highest vote count after adding Gaussian noise n_samples times.

pub fn certify<F: Fn(&[f64]) -> usize>( &self, x: &[f64], base_classifier: &F, ) -> (usize, f64)

Certify: return the L2 radius for which the top-class prediction is guaranteed to hold. Returns 0.0 if not certifiable.

Trait Implementations

impl Clone for RandomizedSmoothingClassifier

fn clone(&self) -> RandomizedSmoothingClassifier

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for RandomizedSmoothingClassifier

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

Formats the value using the given formatter.