Attempting dimensionality reduction on the Census-Income (KDD) Data Set.

The dataset is a mixed dataset with continuous and categorical features.

PCA works fine for continuous variables, reduced down to 2 PCA variables.

[0.84686919 0.14737952]

But having difficulties with categorical variables and cont/cat together.

There are 2 options that are not working out.

  1. multiple correspondence analysis (MCA - for categorical variables)
[0.022592513816454877, 0.022074753604733162, 0.01779832534650477, 0.01644525961669972, 0.013596646212654651]
  1. factor analysis of mixed data (FAMD - for continuous and categorical variables).

Will add these later but they are also very low - not picking up variance in data

Any ideas/tips ?


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.