由历史数据推测未来趋势的众多方法中较突出的有:时间序列法、最小平方法、指数平滑法、回归分析和相关分析。
Prominent among the various techniques that can help to extrapolate past date into future trends are the following:time series, least squares method, exponential smoothing, regression and correlation.
本文在对分位数回归的国内外研究现状进行综述后,介绍了分位数回归的模型和实现方法,并将它与最小平方法、最小一乘法进行了比较。
After summarizing research actuality of quantile regression inside and outside our country, this article introduces this ideal model and realization method, and compares it with OLS and LAD.
我们所提出的方法在整个学习架构上要比强健式支援向量机网路与权重式最小平方支援向量机回归法更简易。
Consequently, the learning mechanism of the proposed approach is much easier than the robust support vector regression networks (RSVRNs) approach and the weighted LS-SVMR approach.
应用推荐