论文部分内容阅读
针对股市指标因子间存在高冗余、非线性等特点,同时开盘价还受政治、经济、投资者心理等因素影响,使得线性方法无法准确地预测开盘价的走势及波动范围.因此,为了提高开盘价预测的准确率,首次提出AP-FIG算法处理收盘价,同时首次在动态改进的粒子群算法(DIPSO)位置更新过程中引入缩减因子,并对其动态调整(DKIPSO),构建基于DKIPSO自动选取SVR参数的开盘价预测模型(DKIPSO-SVR).仿真结果表明,相比于DIPSO-SVR,DKIPSO-SVR不仅提高了泛化能力,还增强了稳健性;对比于GS-SVR,DKIPSO-SVR不仅误差低、拟合效果好,同时还将精度提高近7个百分点.最重要的是,利用AP-FIG算法处理收盘价的预测结果优于其他隶属度方程,使得AP-FIG和DKIPSO-SVR组合模型在股市开盘价预测方面具有更加广泛的应用价值.
Due to the high redundancy and non-linearity between the indexes of the stock market, the opening price is also affected by such factors as politics, economy and investor psychology, which makes the linear method unable to predict the trend and fluctuation range of the opening price accurately. It is the first time that the AP-FIG algorithm is proposed to process the closing price. At the same time, the reduction factor is introduced for the first time in the dynamically updated Particle Swarm Optimization (DIPSO) position updating process, and its dynamic adjustment (DKIPSO) (DKIPSO-SVR) is selected.The simulation results show that compared with DIPSO-SVR, DKIPSO-SVR not only enhances generalization ability but also enhances robustness.Compared with GSK-SVR, DKIPSO-SVR Not only the error is low, the fitting effect is good, but also the accuracy is increased by nearly 7% .Most importantly, the forecast result of using the AP-FIG algorithm to process the closing price is superior to other membership degree equations, so that the AP-FIG and the DKIPSO-SVR Portfolio model in the stock market opening price forecast has a more extensive application value.