What data science technique uses the top $k$ eigenvectors of a covariance matrix to reduce dimensionality?
Show answer
Principal Component Analysis
PCA (Principal Component Analysis) identifies the directions (principal components) along which the variance of the data is maximized. It is mathematically equivalent to the singular value decomposition.
Also asked as:
- Which dimensionality reduction method identifies the directions of maximum variance using a covariance matrix's eigenvectors?