Hi Danny, thanks for your response and feedback. In response to your question, since the selected features are highly correlated, one could replace the 4 features with just a single feature or a couple of features. The proper way to do this is to transform your feature space into a space where the correlation between features is removed. This can be accomplished using a technique called Principal Component Analysis (PCA). You may find an example of this procedure here: Machine Learning: Dimensionality Reduction via Principal Component Analysis.
Thanks again for your response.