摘要
文章针对GARCH模型参数传统估计方法的不足,提出了利用量子粒子群算法的改进算法,并利用此算法实证建立了美国证券市场道琼斯指数收益的GARCH模型,更加精确地动态度量了证券市场收益序列的条件"异方差",并且和基本粒子群算法及其两种改进算法的实验结果进行了比较,最后对指数进行了走势预测.
In this paper, quantum-behaved particle swarm optimization algorithm is developed for some serious disadvantages of traditional estimating methods of GARCH, and the GARCH model for Dow-Jones Average stock return are established empirically, and the results is compared with particle swarm optimization and two improved algorithms, finally forecast of the return is given.
出处
《数值计算与计算机应用》
CSCD
2007年第4期260-266,共7页
Journal on Numerical Methods and Computer Applications
基金
国家自然科学基金(60474030)
关键词
QPSO算法
GARCH模型
异方差
惯性权重法
压缩因子法
QPSO algorithm, GARCH model, heteroskedasticity, inertia weight, con-str^ction factor algorithm
