Question d’entretien chez Amazon

How is variance calculated in a PCA?

Réponses aux questions d'entretien

Utilisateur anonyme

4 juin 2017

For each principle component, the variances should be a linear combination of the variance of the variables that contracted this principle component.

2

Utilisateur anonyme

19 mars 2019

Mathematically, Principal components are a linear combination of eigenvectors and all observations. So in some sense, by multiplying eigenvectors with observations of predictors we implicitly capture the variance.

Utilisateur anonyme

21 mars 2017

How can you sample data from a large dataset?

2