Math — Principal Component Analysis
Dimensionality reduction by variance.
Rendering…
Make it your own.
\Sigma = \frac{1}{n}X^\top X
\Sigma\,\mathbf{v}_i = \lambda_i \mathbf{v}_i
\text{components = leading eigenvectors}
\text{variance explained} = \frac{\lambda_i}{\sum_j \lambda_j}