[−][src]Trait liblinear::LibLinearProblem
Represents a one-to-one mapping of source features to target values.
Source features are represented as sparse vectors of real numbers. Target values are either integers (in classification) or real numbers (in regression).
Required methods
fn source_features(&self) -> &[Vec<FeatureNode>]
The feature vectors of each training instance.
fn target_values(&self) -> &[f64]
Target labels/values of each training instance.
fn bias(&self) -> f64
Bias of the input data.