神经网络在股票型基金业绩持续性判定中的运用
首发时间:2018-05-24
摘要:本文引入神经网络算法,对2014年10月至2017年10月间我国股票型基金业绩数据进行其持续性实证测算。其结果发现,我国股票型基金的持续性在不同时期有一定波动,其波动状态在90%置信度下占13%左右;进行足够学习次数的神经网络算法测算出具有持续性的基金,在考察期内对其业绩是否存在持续性的检测准确率达到70.37%,显著高于传统三种方法测算具有持续性的基金的检测准确率。因此本文认为,神经网络算法能较好的解决现行基金持续性测算方法中的不足,能够提高基金持续性预测的准确度和优化预测的数据结构。
For information in English, please click here
The application of neural network in the continuous determination of stock fund banks
Abstract:This paper introduces a neural network algorithm to conduct continuous empirical measurement of the performance data of China\'s equity funds from October 2014 to October 2017. As a result, it has been found that the sustainability of China\'s equity funds fluctuates at different times, with its volatility at about 90% confidence level of about 13%; a neural network algorithm that conducts enough learning times to calculate a sustainable fund, is inspected The accuracy of the continuous detection of the performance of the company during the period reached 70.37%, which was significantly higher than the traditional three-method measure of the sustainability of the fund\'s detection accuracy. Therefore, this paper believes that the neural network algorithm can better solve the shortcomings of the current fund sustainability measurement methods, can improve the accuracy of the fund\'s continuous forecasting and optimize the forecast data structure.
Keywords: Equity funds Fund performance sustainability Neural network algorithm
基金:
引用
No.****
动态公开评议
共计0人参与
勘误表
神经网络在股票型基金业绩持续性判定中的运用
评论
全部评论0/1000