论文部分内容阅读
针对长期QT心电数据分析中数据量大,传统聚类分析算法在计算快速性和精度性上无法满足要求的问题,提出一种多智能体序列k均值奇异值分解的聚类算法,对QT数据库进行模拟分析。首先,针对数据量大的问题,基于多智能体结构,设计并行化的聚类分析框架,并给出各智能体间的通讯协议;其次,为提高聚类精度,结合奇异值分解理论,对k均值聚类算法进行改进,通过较小随机线性过程对高分辨率的数据进行推断,降低计算复杂度的同时,提高算法的聚类精度;最后,通过在QT数据库上的仿真实验显示,所提算法具有满足要求的实时性和高精度性。
Aiming at the problem of large amount of data in long-term QT ECG data analysis and the traditional clustering analysis algorithm unable to meet the requirements of computational speed and accuracy, a clustering algorithm based on k-means singular value decomposition of multi-agent sequence is proposed. QT Database for simulation analysis. Firstly, in order to solve the problem of large amount of data, a clustering analysis framework based on multi-agent structure and parallelism is designed and the communication protocol among different agents is given. Secondly, in order to improve the clustering accuracy, combined with the theory of singular value decomposition k-means clustering algorithm is improved, the high-resolution data is deduced by a smaller random linear process to reduce the computational complexity and improve the clustering accuracy of the algorithm. Finally, through the simulation experiments on the QT database, The proposed algorithm has to meet the requirements of real-time and high-precision.