Kernel method based on non-linear coherent states in quantum feature space

Abstract In this paper, by mapping datasets to a set of non-linear coherent states, the process encoding inputs in quantum states as feature map is re-interpreted. As result fact that radial basis function recovered when data mapped complex Hilbert state represented can be considered natural generalisation associated kernels. an example kernels based on we propose kernel functions generalized hypergeometric functions, orthogonal polynomial functions. The suggested implemented with support vector machine (SVM) two well known (make_circles, and make_moons) outperforms baselines, even level noise high. addition, study impact geometrical properties space, obtained SVM classification task, considering Fubini–Study metric states.