Struct FittedSparseGP 

pub struct FittedSparseGP<K> {
    pub inducing_points: Array2<f64>,
    pub kernel: K,
    pub approximation: SparseApproximation,
    pub alpha: Array1<f64>,
    pub k_mm_inv: Array2<f64>,
    pub noise_variance: f64,
    pub variational_params: Option<VariationalParams>,
}

Fitted sparse Gaussian Process with learned parameters

Fields

inducing_points: Array2<f64>

Inducing point locations

kernel: K

Kernel function with learned parameters

approximation: SparseApproximation

Sparse approximation method used

alpha: Array1<f64>

Precomputed alpha coefficients

k_mm_inv: Array2<f64>

Inverse of K_mm (inducing point kernel matrix)

noise_variance: f64

Noise variance

variational_params: Option<VariationalParams>

Variational parameters (if using VFE)

Implementations

impl<K: SparseKernel> FittedSparseGP<K>

pub fn predict_with_variance( &self, x: &Array2<f64>, ) -> Result<(Array1<f64>, Array1<f64>)>

Predict with uncertainty quantification

pub fn predict_scalable( &self, x: &Array2<f64>, method: ScalableInferenceMethod, ) -> Result<Array1<f64>>

Scalable prediction using different inference methods

pub fn log_marginal_likelihood(&self) -> Result<f64>

Compute log marginal likelihood for model selection

pub fn sample_posterior( &self, x: &Array2<f64>, num_samples: usize, ) -> Result<Array2<f64>>

Sample from the posterior predictive distribution

pub fn acquisition_function( &self, x: &Array2<f64>, acquisition_type: &str, ) -> Result<Array1<f64>>

Compute acquisition function for active learning

Trait Implementations

impl<K: Clone> Clone for FittedSparseGP<K>

fn clone(&self) -> FittedSparseGP<K>

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl<K: Debug> Debug for FittedSparseGP<K>

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

Formats the value using the given formatter.

impl<K: SparseKernel> Predict<ArrayBase<OwnedRepr<f64>, Dim<[usize; 2]>>, ArrayBase<OwnedRepr<f64>, Dim<[usize; 1]>>> for FittedSparseGP<K>

Prediction implementation for fitted sparse GP

fn predict(&self, x: &Array2<f64>) -> Result<Array1<f64>>

Make predictions on the provided data

fn predict_with_uncertainty( &self, x: &X, ) -> Result<(Output, UncertaintyMeasure), SklearsError>

Make predictions with confidence intervals