Module omp 

Orthogonal Matching Pursuit (OMP).

This module provides OrthogonalMatchingPursuit, a greedy feature selection algorithm that iteratively selects the feature most correlated with the current residual, adds it to a support set, solves OLS on the support, and updates the residual. The process repeats until the desired number of non-zero coefficients is reached or the residual tolerance is met.

Examples

use ferrolearn_linear::OrthogonalMatchingPursuit;
use ferrolearn_core::{Fit, Predict};
use ndarray::{array, Array1, Array2};

let x = Array2::from_shape_vec((5, 3), vec![
    1.0, 0.0, 0.0,
    2.0, 0.1, 0.0,
    3.0, 0.0, 0.1,
    4.0, 0.1, 0.0,
    5.0, 0.0, 0.1,
]).unwrap();
let y = array![2.0, 4.0, 6.0, 8.0, 10.0];

let model = OrthogonalMatchingPursuit::<f64>::new().with_n_nonzero_coefs(1);
let fitted = model.fit(&x, &y).unwrap();
let preds = fitted.predict(&x).unwrap();
assert_eq!(preds.len(), 5);

Structs

FittedOMP

Fitted Orthogonal Matching Pursuit model.

OrthogonalMatchingPursuit

Orthogonal Matching Pursuit.