| 应用不等权重支持向量机预测人民币汇率的变动 |
| |
| 引用本文： | 徐成贤,陈静,王昭. 应用不等权重支持向量机预测人民币汇率的变动[J]. 统计与信息论坛, 2010, 25(10): 45-48 |
| |
| 作者姓名： | 徐成贤 陈静 王昭 |
| |
| 作者单位： | [1]杭州师范大学杭州国际服务工程学院,浙江杭州310012 [2]西安交通大学经济与金融学院,陕西西安710061 [3]西安交通大学理学院,陕西西安710049 |
| |
| 基金项目： | 国家自然科学基金，杭州师范大学科研项目 |
| |
| 摘 要： | 基于金融时间序列的近期数据对未来的影响会大于早期数据,对应用于金融时间序列预测的支持向量机方法进行改进,给出了不等权重支持向量机方法（USVM）及其多项式光滑化处理。将不等权重支持向量机方法应用于训练样本集的子集确定预测模型,实证分析表明USVM算法预测是有效的。
|
| 关 键 词： | 支持向量机 遗传算法 多项式光滑 汇率预测 |
Apply Support Vector Machine with Unequal Weights to Forecast CNY Exchange Rate Movement Direction |
| |
| XU Chen-xian,CHEN Jing,WANG Zhao. Apply Support Vector Machine with Unequal Weights to Forecast CNY Exchange Rate Movement Direction[J]. Statistics & Information Tribune, 2010, 25(10): 45-48 |
| |
| Authors: | XU Chen-xian CHEN Jing WANG Zhao |
| |
| Affiliation: | 1.Hangzhou Institute of International Service,HangZhou Normal University,HangZhou 310012,China;2.a.School of Economics＆ Finance,Xi＇an Jiaotong University,Xi＇an 710061,China;b.School of Science,Xi＇an Jiaotong University,Xi＇an 710049,China) |
| |
| Abstract: | A support vector machine with unequal weights（USVM） is given based on the fact that the latest data gives more effects on the future than the early data does in financial time series forecasting.The problem generated from USVM is converted into a non-smooth unconstrained optimization problem and the resulting problem is then smoothed using polynomials.Empirical analyses on USVM are made in forecasting the movement directions of foreign exchange rates.And the result shows the efficiency of USVM. |
| |
| Keywords: | support vector machine polynomial smoothing foreign exchange-rate forecast |
| 本文献已被 维普 万方数据 等数据库收录！ |
|
