Struct ADMMSVM 

pub struct ADMMSVM { /* private fields */ }

ADMM (Alternating Direction Method of Multipliers) SVM Solver

ADMM is a distributed optimization algorithm that decomposes the SVM problem into smaller subproblems that can be solved in parallel. It’s particularly effective for large-scale and distributed SVM training.

The algorithm solves the consensus problem: minimize f(x) + g(z) subject to Ax + Bz = c

For SVM, this becomes: minimize (1/2)||w||² + C∑ξᵢ subject to yᵢ(wᵀφ(xᵢ) + b) ≥ 1 - ξᵢ, ξᵢ ≥ 0

Reference: Boyd, S. et al. (2011). Distributed optimization and statistical learning via the alternating direction method of multipliers.

Implementations

impl ADMMSVM

pub fn new(config: AdvancedOptimizationConfig) -> Self

Create a new ADMM SVM solver

pub fn fit( &mut self, x: &Array2<f64>, y: &Array1<f64>, ) -> Result<OptimizationResult>

Fit the SVM using ADMM optimization

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

Predict using the fitted model

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

Calculate decision function values

Trait Implementations

impl Clone for ADMMSVM

fn clone(&self) -> ADMMSVM

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for ADMMSVM

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

Formats the value using the given formatter.

impl Default for ADMMSVM

fn default() -> Self

Create a new ADMM SVM solver with default configuration