摘要
提出一种基于最小二乘支持向量机(LS-SVM)的福房指数预测方法。采用感知机核函数、多项式核函数和高斯核函数进行仿真模拟,经过参数选优建立了精度较高的福房指数预测模型。预测结果表明,利用LS-SVM模型进行预测具有误差小、拟合程度高等优点,可适用于房地产价格指数的预测。
A method of forecasting Fuzhou real estate price index based on least squares support vector machines （LS-SVM） was proposed. The kernel functions （perceptron function, polynomial and Gaussian kernel functions） were used in simulation. A relatively high precision model of Fuzhou real estate price index forecast was built by optimizing parameters. The forecasting results show that the model is of smaller error and higher fitting and is applicable in real estate price index forecast.
出处
《福建工程学院学报》
CAS
2009年第4期393-396,共4页
Journal of Fujian University of Technology
