You've applied PCA but the variance explained by the first few components is very low. What could be the underlying issue and how might you remedy it?

  • The data has no variance, so PCA is not applicable
  • The data is not centered, so you should center it before applying PCA
  • The data is too complex for PCA, so you should switch algorithms
  • The eigenvalues have been miscalculated and you should recalculate them
If the variance explained by the first few components is very low, it may be because the data is not centered. Centering the data by subtracting the mean is a necessary preprocessing step for PCA.
Add your answer
Loading...

Leave a comment

Your email address will not be published. Required fields are marked *