Optimizing Variational Quantum Neural Networks Based on Collective Intelligence

• DOI: 10.3390/math12111627
• Publication Date: 2024-05-23T07:49:03Z
• Authors: Zitong Li; Tailong Xiao; Xiaoyang Deng; Guihua Zeng; Weimin Li
• Keywords: Collective Intelligence

Quantum machine learning stands out as one of the most promising applications quantum computing, widely believed to possess potential advantages. In era noisy intermediate-scale quantum, scale and quality computers are limited, algorithms based on fault-tolerant computing paradigms cannot be experimentally verified in short term. The variational algorithm design paradigm can better adapt practical characteristics hardware is currently solutions. However, algorithms, due their highly entangled nature, encounter phenomenon known “barren plateau” during optimization training processes, making effective challenging. This paper addresses this challenging issue by researching a neural network method collective intelligence algorithms. aim overcome difficulties encountered traditional methods such gradient descent. We study two typical using networks: random 2D Hamiltonian ground state solving phase recognition. find that shows compared solution accuracy energy classification enhanced, iterations also reduced. highlight has great tackling