小波支持向量机理论及其在股指预测中的应用

发布时间：2018-03-18 10:13

本文选题：支持向量机　切入点：时间序列预测　出处：《西南财经大学》2013年硕士论文　论文类型：学位论文

【摘要】：支持向量机理论是近几年发展起来,专门处理小样本的条件下分类与预测问题的新方法,源自于机器学,同时广泛应用于时间序列预测理论和金融市场预测等方面,能够有效解决小样本、非线性、高维数和局部极小值等问题。然而,标准支持向量机假设的样本分布过于理想,处理复杂的实际问题时难免遇到困难。因此,本文以解决实际问题为线索,根据对时间序列数据预测的要求,以提高计算效率,降低优化问题复杂度为前提,探索一些极端条件下支持向量机的新理论、新方法。本文在文献[1]的基础上作了一些工作。文献[1]中阐述了单变量morlet小波应用到支持向量机理论的模型并利用了小波支持向量机方法来对金融市场数据进行预测,与文献[1]不同的是,本文中将探索基于morlet小波支持向量机与多变量预测技术相结合的方法并将其应用到对股票指数的预测中。
[Abstract]:Support vector machine (SVM) theory is a new method developed in recent years to deal with classification and prediction problems under the condition of small samples. It originates from machine science and is widely used in time series forecasting theory and financial market forecasting. It can effectively solve the problems of small sample, nonlinear, high dimension and local minimum. However, the sample distribution of the assumption of standard support vector machine is too ideal, so it is difficult to deal with complex practical problems. Based on the clue of solving practical problems and the requirement of time series data prediction, this paper explores new theories and methods of support vector machine under some extreme conditions on the premise of improving computational efficiency and reducing the complexity of optimization problems. In this paper, some work is done on the basis of reference [1]. In reference [1], the model of applying univariate morlet wavelet to support vector machine theory is described and the wavelet support vector machine method is used to predict the financial market data. Different from the reference [1], this paper will explore the method of combining morlet wavelet support vector machine with multivariable prediction technology and apply it to the prediction of stock index.
【学位授予单位】：西南财经大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：F830.91;TP181

【参考文献】