Keyboard shortcuts

Press ← or → to navigate between chapters

Press S or / to search in the book

Press ? to show this help

Press Esc to hide this help

Multi-class & multi-output

solow-multi covers one-vs-rest, one-vs-one, output-code, multi-output regressor and classifier, and chain estimators that model target correlations.

Multi-class

EstimatorWhat it does
OneVsRestClassifierFit one binary classifier per class against the rest.
OneVsOneClassifierFit one binary classifier per class pair and vote.
OutputCodeClassifierError-correcting output codes with configurable code size and random codebook.

Multi-output

EstimatorWhat it does
MultiOutputRegressorFit an independent regressor per output.
MultiOutputClassifierFit an independent classifier per output.
RegressorChainSequential regressor chain where each target uses previous target predictions.
ClassifierChainSequential classifier chain, modelling label correlations.

Example

#![allow(unused)]
fn main() {
use solow_multi::OneVsRestClassifier;

let ovr = OneVsRestClassifier::new(base_binary_classifier).fit(&x, &y)?;
let proba = ovr.predict_proba(&x_test)?;
Ok::<_, solow_core::Error>(())
}