Struct GraphSemiSupervisedSVM 

pub struct GraphSemiSupervisedSVM {
    pub c: f64,
    pub gamma_a: f64,
    pub gamma_i: f64,
    pub kernel: String,
    pub graph_kernel: String,
    pub n_neighbors: usize,
    pub sigma: f64,
    pub max_iter: usize,
    pub tol: f64,
    pub verbose: bool,
}

Graph-based Semi-supervised Support Vector Machine

This implementation combines SVM with graph-based label propagation, using the graph structure to regularize the learning process and propagate information from labeled to unlabeled examples.

Parameters

• C - Regularization parameter for SVM (default: 1.0)
• gamma_a - Graph regularization parameter (default: 0.1)
• gamma_i - Intrinsic regularization parameter (default: 0.01)
• kernel - Kernel function for SVM (‘rbf’, ‘linear’, default: ‘rbf’)
• graph_kernel - Kernel for graph construction (‘rbf’, ‘knn’, default: ‘rbf’)
• n_neighbors - Number of neighbors for k-NN graph (default: 10)
• sigma - Bandwidth for RBF kernel (default: 1.0)
• max_iter - Maximum number of iterations (default: 1000)
• tol - Tolerance for convergence (default: 1e-4)
• verbose - Enable verbose output (default: false)

Example

use sklears_svm::GraphSemiSupervisedSVM;
use sklears_core::traits::{Predict};
use scirs2_core::ndarray::array;

let X_labeled = array![[1.0, 2.0], [2.0, 3.0]];
let y_labeled = array![0, 1];
let X_unlabeled = array![[1.5, 2.5], [2.5, 3.5]];

let model = GraphSemiSupervisedSVM::new()
    .with_c(1.0)
    .with_gamma_a(0.1);

let trained_model = model
    .fit_semi_supervised(&X_labeled, &y_labeled, &X_unlabeled)
    .expect("GraphSemiSupervisedSVM fit_semi_supervised should succeed on valid input");
let predictions = trained_model.predict(&X_labeled).expect("GraphSemiSupervisedSVM predict should succeed on valid input");

Fields

c: f64

SVM regularization parameter

gamma_a: f64

Graph regularization parameter

gamma_i: f64

Intrinsic regularization parameter

kernel: String

Kernel function for SVM

graph_kernel: String

Kernel function for graph construction

n_neighbors: usize

Number of neighbors for k-NN graph

sigma: f64

Bandwidth for RBF kernel

max_iter: usize

Maximum number of iterations

tol: f64

Tolerance for convergence

verbose: bool

Verbose output

Implementations

impl GraphSemiSupervisedSVM

pub fn new() -> Self

Create a new GraphSemiSupervisedSVM with default parameters

pub fn with_c(self, c: f64) -> Self

Set the SVM regularization parameter C

pub fn with_gamma_a(self, gamma_a: f64) -> Self

Set the graph regularization parameter

pub fn with_gamma_i(self, gamma_i: f64) -> Self

Set the intrinsic regularization parameter

pub fn with_kernel(self, kernel: &str) -> Self

Set the SVM kernel function

pub fn with_graph_kernel(self, graph_kernel: &str) -> Self

Set the graph construction kernel

pub fn with_n_neighbors(self, n_neighbors: usize) -> Self

Set the number of neighbors for k-NN graph

pub fn with_sigma(self, sigma: f64) -> Self

Set the RBF kernel bandwidth

pub fn with_max_iter(self, max_iter: usize) -> Self

Set the maximum number of iterations

pub fn with_tol(self, tol: f64) -> Self

Set the tolerance for convergence

pub fn with_verbose(self, verbose: bool) -> Self

Set verbose output

pub fn fit_semi_supervised( self, x_labeled: &Array2<f64>, y_labeled: &Array1<i32>, x_unlabeled: &Array2<f64>, ) -> Result<TrainedGraphSemiSupervisedSVM>

Train graph-based semi-supervised SVM

Trait Implementations

impl Clone for GraphSemiSupervisedSVM

fn clone(&self) -> GraphSemiSupervisedSVM

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for GraphSemiSupervisedSVM

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

Formats the value using the given formatter.

impl Default for GraphSemiSupervisedSVM

fn default() -> Self

Returns the “default value” for a type.

impl Estimator for GraphSemiSupervisedSVM

type Config = GraphSemiSupervisedSVM

Configuration type for the estimator

type Error = SklearsError

Error type for the estimator

type Float = f64

The numeric type used by this estimator

fn config(&self) -> &Self::Config

Get estimator configuration

fn validate_config(&self) -> Result<(), SklearsError>

Validate estimator configuration with detailed error context

fn check_compatibility( &self, n_samples: usize, n_features: usize, ) -> Result<(), SklearsError>

Check if estimator is compatible with given data dimensions

fn metadata(&self) -> EstimatorMetadata

Get estimator metadata