摘要
对作为数据挖掘模型预测项的股票波段数值特征进行了研究.首先采用了二次平滑模型和有限自动机模型,根据股票峰(谷)点的概念实现对波段状态的初步提取.然后根据波段对象属性的规范化表示,给出了小波段与盘整波段的规则设定,用有限自动机模型实现小波段合并与盘整波段的提取,最终得到了过滤后的波段数值特征.
The numerical characteristics of stock wave were studied based on pattern recognition. The quadratie smoothing model and basic limited automatic theories were adopted to extract the stock state according to the theory of stock peak-valley point. Daily transaction state of stock was realized to preliminary extraction. After standard expression of wave band, rule-making of the whole wave band and filtering of partial wave band, the numerical characteristics of stock wave were obtained finally.
出处
《成都大学学报(自然科学版)》
2012年第2期161-163,共3页
Journal of Chengdu University(Natural Science Edition)
关键词
股市行情
二次平滑
有限自动机
波段对象
stock wave
quadratic smoothing model
finite automation
wave band