Principal Component Analysis

Let's say I am predicting a binary outcome with a mix of numeric and categorical independent variables. If I know the numeric variables are correlated I will use PCA to reduce them to smaller principal components. Then, can I use the PC's as predictors along with the unmanipulated categorical predictors? Thanks.


Less is more. Stay pure. Stay poor.
Probably, not my area, but I think that is what people do. Why are some variables correlated is what I would be asking myself. I haven't done a PCA before, but did factor analysis once a long, long time ago.
So, I'm building a machine learning algorithm to predict repairs for a vehicle. I suspect a lot of the the numeric attributes like vehicle age, mileage, and vehicle value ($) amongst other variables are correlated.