What is a covariance matrix in PCA?

Updated May 16, 2026

Short answer

A covariance matrix captures relationships between features in PCA.

Deep explanation

It shows how variables vary together. PCA uses eigenvectors of the covariance matrix to find principal components.

Real-world example

Understanding how stock prices move together in finance.

Common mistakes

  • Ignoring scaling before computing covariance.

Follow-up questions

  • What does positive covariance mean?
  • Why is covariance important in PCA?

More Dimensionality Reduction interview questions

View all →