基于AGSO-LSSVM的热点话题预测模型

Hot topic prediction model based on AGSO-LSSVM

针对网络热点话题的时变性、混沌性,为了进一步提高网络热点话题的预测精度,提出一种人工萤火虫算法(artificial glowworm swarm optimization,AGSO)优化最小二乘支持向量机(least squares support vector machine,LSSVM)的网络热点话题预测模型(AGSO-LSSVM)。模型收集网络热点话题数据,采用互信息法和CAO法选择最优延迟时间和嵌入维数,采用优延迟时间和嵌入维数重构网络热点话题数据学习样,并输入到最小二乘支持向量机中训练,进而采用人工萤火虫算法优化最小二乘支持向量机参数建立网络热点话题预测模型,采用仿真实验测试其性能。实验结果表明,相对于其他网络热点话题预测模型,该模型可以对网络热点话题的变化特点进行拟合,进一步提高网络热点话题的预测准确性。

This paper proposes a hot topic chaotic prediction model based on glowworm swarm algorithm optimizing least squares support vector machine（AGSO-LSSVM） because hot topic has time-varying and chaotic characteristics. Firstly, the optimal time delay and embedding dimension are determined by using C-C method and CAO method, and the time series of hot are reconstructed, and then least squares support vector is used to build hot topic prediction model, and artificial glow- worm swarm optimization algorithm is used to optimize the parameters of least squares support vector machine. Finally, the simulation experiment is carried out to test the performance of model. The resuhs show that, compared with the other mod- els, the proposed model has improved the prediction accuracy of hot topic.