Variable Hyperparameterized Gaussian Kernel using Displaced Squeezed Vacuum State

Abstract

There are schemes for realizing different types of kernels by quantum states of light. It is particularly interesting to realize the Gaussian kernel due to its wider applicability. A multimode coherent state can generate the Gaussian kernel with a constant value of hyperparameter. This constant hyperparameter has limited the application of the Gaussian kernel when it is applied to complex learning problems. We realize the variable hyperparameterized Gaussian kernel with a multimode-displaced squeezed vacuum state. The learning capacity of this kernel is tested with the support vector machines over some synthesized data sets as well as public benchmark data sets. We establish that the proposed variable hyperparameterized Gaussian kernel offers better accuracy over the constant Gaussian kernel.