Table 1 Principles of the dimensionality reduction techniques used in the classification and prediction algorithms.

From: Translational utility of a hierarchical classification strategy in biomolecular data analytics

Method

Method Abbrev.

Components derivation

Principal Component Analysis

PCA

Maximizes overall dataset variance without considering between-class variance

Partial Least Squares

PLS

Maximizes between-class variance without considering within-class variance

Maximum Margin Criterion

MMC

Maximizes between-class variance, while minimizing within-class variance

Linear Discriminant Analysis

LDA

Maximizes ratio of between- and within-class variation while the number of samples is greater than the number of variables

Support Vector Machines

SVM

Maximizes the margin of separation between the classes

  1. Methods, their respective abbreviation and a descriptive derivation of their components to obtain a reduced dimensionality space. PCA and LDA were used in combination with each other or with other methods to achieve the combinatory methods: PCA-LDA and MMC-LDA.
Back to article page