Abstract:
A lot of work shows that quantum computing can speed up supervised and unsupervised learning algorithms, even improving the performance of these algorithms. Usually the following methods are used: based on quantum theory, it encodes classical information into quantum state, and then represents all of samples in dataset by one or several quantum states; It quantifies the similarity of two samples by measuring the distance between the quantum states through a quantum algorithm. The theoretical and simulation results show that quantum computation can accelerate classical machine learning algorithms. Finally, it summarizes the advantages of quantum machine learning technology and the existing problems, and looks forward to the future development trend of this field.