Using continuous and categorical features together for classification

I am trying a deep neural network for classification in keras for the dataset uploaded with the post.
I am wondering how can I use both continuous and categorical features together to further improve my accuracy, the dataset has 150000 entries. Should I one-hot encode the label-encoded features, if yes what is the proper way to do it in keras?

Description of the data:-

cont_x = Think of a connection as a sequence of TCP packets starting and ending at some well defined times, between which data flows to and from a source IP address to a target IP address under some well-defined protocol.

cat_x = under every column, categories have been label encoded (a = 1, b = 2 etc)

Target = In total, there are 40 type of attacks to which their network is vulnerable to. But, 3 of them cause the maximum damage.Predict the type of attack(s).

data_train_sample1.pdf (15.6 KB)

If you decide to switch to Tensorflow, there’s this great article that gives you an overview on how to do that.