Nominal Attributes in LibSVM

When creating a libsvm training file, how do you differentiate between a nominal attribute verses a numeric attribute? I'm trying to encode certain nominal attributes as integers, but I want to ensure libsvm doesn't misinterpret them as numeric values. Unfortunately, libsvm's site seems to have very little documentation. Pentaho's docs seem to imply libsvm makes this distinction, but I'm开发者_开发技巧 still not clear how it's made.

Don't do this I'm trying to encode certain nominal attributes as integers.

Rather, use a separate binary feature for each value of each nominal attribute.

The way SVMs are formulated, all attributes/features are numeric and class labels are nominal. Nominal attributes are essentially faked by using mutually exclusive binary features.

I think you cant do that in libsvm, weka or SVM-light. One approach that you could use is to use something like a decision tree for your nominal attributes and svm or any distance based classifier for your numeric attributes and then combine the results. I hope it helps.