摘要
微博信息传播预测研究已成为广大学者研究的重点,而微博转发是其中一个关键机制。本文结合传统的传染病模型,融入外来用户影响因子,得到改进的微博预测SIRE模型。通过对比SISe模型,SIRE模型能够更好地拟合和预测微博转发规律。从算法性能上通过对比和声算法、粒子群算法、遗传算法,结果表明,采用和声算法进行微博转发预测模型参数寻优,其性能最佳,并能很好地表征微博传播趋势。
The research on micro blog communication prediction was a hot spot. Micro blog forwarding was one of the key mechanisms. On the basis of traditional infectious diseases model, an improved SIRE model was proposed through integrating into the foreign user impact factor. Comparing with the SISe model, SIRE model could be used to fit well and forecast the micro blog forwarding rule. Comparing with Harmony Algorithm, the Particle Swarm Algorithm and Genetic Algorithm, Harmony Algorithm was used for parameters optimization of micro blog forwarding prediction model, its performance was the best, it could characterize the tread of micro blog communication.
出处
《铁路计算机应用》
2016年第2期12-15,共4页
Railway Computer Application
关键词
传染病模型
转发行为
和声算法
参数优化
infectious disease model
forwarding behavior
Harmony Algorithm
parameters optimization
