Developer Guide for Intel® Data Analytics Acceleration Library 2018
Given the results of the PCA algorithm, data set E = (e i ), i = 1, ..., p of eigenvalues in decreasing order, full number of principal components p and reduced number of components p r ≤ p the problem is to evaluate the explained variances radio and noise variance.
The metrics are computed given the input data meets the following requirements:
The PCA quality metrics include:
The library uses the following quality metrics:
Quality Metric |
Definition |
---|---|
Explained variance |
Eigenvalues: |
Explained variance ratios |
![]() |
Noise variance |
![]() p r - number of principal components, p - number of features in the data set |