Quantum Expectation-Maximization Algorithm

Quantum Expectation-Maximization Algorithm
复制标题

DOI:
10.1103/physreva.101.012326
复制
发表时间:
2019-08
期刊:
ArXiv
影响因子:
--
通讯作者:
Hideyuki Miyahara;K. Aihara;W. Lechner
Hideyuki Miyahara;K. Aihara;W. Lechner
中科院分区:
其他
文献类型:
--
作者:
Hideyuki Miyahara;K. Aihara;W. Lechner

文献摘要

相似文献

聚类算法是机器学习应用的基石。最近,Kerenidis, Landman, Luongo和Prakash提出了一种基于k -means算法的聚类量子算法。基于他们的工作,我们提出了高斯混合模型(GMMs)的量子期望最大化算法。证明了该算法的鲁棒性和量子加速性。我们还在数值上展示了GMM在非平凡聚类数据上优于$k$-means算法的优势。
Clustering algorithms are a cornerstone of machine learning applications. Recently, a quantum algorithm for clustering based on the $k$-means algorithm has been proposed by Kerenidis, Landman, Luongo, and Prakash. Based on their work, we propose a quantum expectation-maximization algorithm for Gaussian mixture models (GMMs). The robustness and quantum speedup of the algorithm are shown. We also show numerically the advantage of GMM over $k$-means algorithm for nontrivial cluster data.