Principal Component Analysis#
Principal Component Analysis (PCA) is a dimensionality reduction technique that aims to transform the feature space by the k principal components that explain the most variance. PCA is used to compress high-dimensional samples down to lower dimensions such that they would retain as much information as possible.
Interfaces: Transformer, Stateful, Persistable
Data Type Compatibility: Continuous only
Parameters#
| # | Name | Default | Type | Description |
|---|---|---|---|---|
| 1 | dimensions | int | The target number of dimensions to project onto. |
Example#
use Rubix\ML\Transformers\PrincipalComponentAnalysis;
$transformer = new PrincipalComponentAnalysis(15);
Additional Methods#
Return the amount of variance that has been preserved by the transformation:
public explainedVar() : ?float
Return the amount of variance lost by discarding the noise components:
public noiseVar() : ?float
Return the percentage of information lost due to the transformation:
public lossiness() : ?float
References#
-
H. Abdi et al. (2010). Principal Component Analysis. ↩
Last update: 2021-03-03