Acceleration techniques for optimization over trained neural network ensembles - 42Papers