摘要： 为描述网络论坛文章数序列的自相似性，使用基于分数差分和ARMA过程的FARIMA模型，对论坛的文章数序列进行建模和分析。FARIMA模型具有描述精度高的特点，但时间复杂度较高。为此，提出?-ARMA模型。通过实验进行仿真分析，从模型的计算复杂度和拟合准确度方面进行对比，证明了?-ARMA模型的有效性。
关键词:
网络论坛,
FARIMA模型,
自相似
Abstract: In order to describe the self-similarity of the online forums article number sequences, the FARIMA process based on fractional differential and the ARMA process is employed to study and analyze the sequences. FARIMA model has the characteristics of high accuracy and time complexity. In order to reduce the time complexity of FARIMA model, ?-ARMA model is proposed. Simulation experiments compare the computational complexity and fitting accuracy, and the results prove the validity of the ?-ARMA model.
Key words:
online forum,
FARIMA model,
self-similarity
中图分类号:
范文嵩, 曾剑平, 张世永. 自相似网络论坛的模型研究[J]. 计算机工程, 2011, 37(6): 260-262.
FAN Wen-Song, CENG Jian-Beng, ZHANG Shi-Yong. Research on Models of Self-similar Online Forum[J]. Computer Engineering, 2011, 37(6): 260-262.