HCPM

Extends the traditional correlation matrix (between numerical data) to also
include binary and ordinal categorical data and computes the p-values for the tests of uncorrelation.

Unsupervised multivariate filter feature selection using the UFS-rHCM or UFS-cHCM algorithms based on the heterogeneous correlation matrix (HCM). The HCM consists of Pearson's correlations between numerical features, polyserial correlations between numerical and ordinal features, and polychoric correlations between ordinal features. Tortora C., Madhvani S., Punzo A. (2025). "Designing unsupervised mixed-type feature selection techniques using the heterogeneous correlation matrix." International Statistical Review <doi:10.1111/insr.70016>. This work was supported by the National Science foundation NSF Grant N 2209974 (Tortora) and by the Italian Ministry of University and Research (MUR) under the PRIN 2022 grant number 2022XRHT8R (CUP: E53D23005950006), as part of ‘The SMILE Project: Statistical Modelling and Inference to Live the Environment’, funded by the European Union – Next Generation EU (Punzo).

Cristina Tortora

hetcorFS

Unsupervised Feature Selection using the Heterogeneous
Correlation Matrix

Antonio Punzo

Shaam Madhvani

HCPM function

<dl><dt>data</dt>
<dd>A data frame. Values of type 'numeric' or 'integer' are treated as numerical.</dd></dl>

Arguments

Heterogeneous correlation and p-value matrices — HCPM

<dl>

<dt>data</dt>
<dd>A data frame. Values of type 'numeric' or 'integer' are treated as numerical.</dd>

</dl>

HCPM: Heterogeneous correlation and p-value matrices

Description

Usage

Value

Arguments

References

Examples