Action Schema Neural Networks Generalized Policies for Stochastic Planning Problems In The Wargaming Domain