Prediction of Energy Gaps in Graphene—Hexagonal Boron Nitride Nanoflakes Using Artificial Neural Networks