Quantum artificial neural network architectures and components

摘要: It is shown by classical simulation and experimentation that quantum artificial neural networks (QUANNs) are more efficient in some cases powerful than (CLANNs) for a variety of experimental tasks. This effect particularly noticeable with larger complex domains. The gain efficiency achieved no generalisation loss most cases. QUANNs also CLANNs, again the tasks examined, terms what network can learn. What more, it appears not all components QUANN architecture need to be these advantages surface. demonstrated fully has advantage over partly may fact produce worse results. Overall, this work provides first insight into expected behaviour individual QUANNs, if when hardware ever built, raises questions about interface between future QUANNs.