Hybrid ANN and SVM

I have a dataset including 10 continues and 5 categorical variables with 5000 samples. I need to run machine learning algorithms for this dataset to be able to use it as a cost function for GA optimization. As ANN fits well for continues data and SVM is very appropriate for categorical data, do you think it is reasonable to split the dataset into two parts of continues and categorical data and then, run ANN and SVM for them separately? if so, is there any consolidated method to aggregate the results into one package so that provided as cost function for GA optimization?

Thank you.