Adapting a Neural Network to efficiently run on a Quantum Circuit

Is there a way to adapt neural networks to take advantage of quantum phenomena such as entanglement and superposition to speed up the learning process?