Stochastic Neighbor Embedding Algorithm Based on Quantum Genetic Algorithm with Gaussian Parameters

Abstract

Stochastic neighbor embedding algorithm is an important nonlinear dimension reduction manifold learning algorithm in the field of big data and machine learning. In the stochastic neighborhood embedding algorithm, it changes the idea of constant distance based on the medium in and, while mapping high-dimensional to low-dimensional, trying to ensure that the distribution probability of each other is consistent. The gradient descent method is often used to solve the problem of minimum divergence, but because the gradient descent method has the disadvantage of easily falling into local optimal values, this article combines SNE with a quantum genetic algorithm and uses the strong uncertainty of the quantum genetic algorithm and high convergence to solve the problem.