摘要
本文采取了一种基于支持向量机的预测方法对公共事业费的支出情况进行了预测。计算结果证明,这种方法与传统的方法相比,具有更好的泛化能力,可以使神经网络方法出现的过拟合、网络结构难以确定等问题得到有效地解决,因而该方法可以作为一种新途径用于研究此类问题。
In this paper, a new forecasting method based on support vector machine is adopted to predict the expenditure of public utilities. The calculation results show that this method, compared with the traditional method, has better generalization ability, can effectively solve over-simulated problem, the network structure difficult to be determined and others which existing in the the neural network method,thus the method can be used as a new approach for the study of such problems.
出处
《价值工程》
2016年第29期9-11,共3页
Value Engineering
关键词
支持向量机
公共事业费
费用预测
support vector machine
public utilities
cost forecast
