Speaker
Description
Quantum Machine Learning is the field that aims to integrate machine learning into quantum computation. In the past years, some works have shown that we can naturally generate one-dimensional Fourier series with a simple supervised quantum learning model. However, there is a lack of explanation of such models for generating Fourier series of larger data-dimension. In this work, we provide a detailed study of different quantum strategies for producing higher-dimensional Fourier series. Besides this, we show that the generalization of the one-dimensional model requires more degrees of freedom than the degrees disposed of in the whole Hilbert space of the circuit model. These results exhibit that we can not fit an arbitrary multi-dimensional function with such models. Nevertheless, the models can be used to approximate functions up to a certain degree. We propose several alternatives to the models, which contribute to the study of multi-feature quantum machine learning algorithms with classical data.
