If data is correlated PCA does not work well. Because of the correlation of variables the effective variance of variables decreases. Hence correlated data when used for PCA does not work well.

In the context of machine learning interview questions, if you are asked about the impact of correlation on Principal Component Analysis (PCA), here is a concise and accurate response:

“Correlation affects PCA in the sense that highly correlated features may result in a less effective dimensionality reduction. PCA works best when features are uncorrelated or weakly correlated because it aims to capture the maximum variance in the data. If features are highly correlated, the principal components may end up capturing redundant information. In such cases, it might be beneficial to consider addressing multicollinearity before applying PCA, for example, by removing or combining correlated features.”