midPCA

What is covariance matrix in PCA?

Updated May 17, 2026

Short answer

It represents pairwise feature relationships used to compute principal components.

Deep explanation

Covariance matrix captures how features vary together. PCA uses it to identify directions of maximum variance by decomposing this matrix into eigenvalues and eigenvectors.

Real-world example

Understanding correlation in financial indicators.

Common mistakes

  • Confusing covariance with correlation.

Follow-up questions

  • What is diagonal of covariance matrix?
  • Why transpose X?

More PCA interview questions

View all →