Time Series Analysis: Forecasting and Control
Time Series Analysis: Forecasting and Control
复制标题
DOI：
10.1080/00401706.1977.10489562
复制
发表时间：
1977-08
期刊：
影响因子：
2.5
通讯作者：
B. Kleiner
中科院分区：
文献类型：
--
作者：
B. Kleiner
这本书提供了一个写得很好的和有组织的短期单变量预测方法的介绍。它主要针对指数平滑模型及其各种扩展。其他方法也有描述，但在文中并不起重要作用。所有的过程都是内生的，只使用要预测的系列的历史。它的主要贡献之一是以一种明确和一致的方式汇集了关于短期预测模型的相当广泛和多样化的文献。全书整体流程流畅，对预测中的标准问题进行了介绍和讨论。这本书由十章和三个附录组成。在详细介绍常用术语的定义（例如，交货时间，预测期等）以及通常的激励材料之后，还有关于回归和指数平滑模型（包括加权直接平滑过程）的章节。接下来是关于预测和预测误差分析的章节，并扩展到根据预测误差的顺序修改平滑参数。本书最后分别讨论了预测中的“Box-Jenkins模型”和贝叶斯方法。附录提供了常用的统计表、本文示例中使用的实际时间序列数据的集合以及两个用于预测的计算机程序。这本书的风格是直接的，因为主题是以一种良好的组织和顺序的方式呈现的。它被设计用于本科预测课程的水平
This book provides a well written and organized introduction to short-term univariate forecasting methods. It is heavily directed toward exponential smoothing models and their various extensions. Other methods are described but do not play an important role in the text. All the procedures are endogenous, using only the history of the series to be forecast. One of its major contributions is bringing together the rather broad and diverse literature on shortterm forecasting models in a clear and consistent manner. The overall flow of the book is smooth, with the standard issues in forecasting introduced and discussed. The book consists of ten chapters and three appendices. Following a fine introduction covering definitions of often used terms (eg, lead time, forecast period, etc.) as well as the usual motivational material, come chapters on regression and exponential smoothing models (including weighted direct smoothing procedures). Next are chapters on forecasting and the analysis of forecast errors with extensions to modifying smoothing parameters based on the sequence of forecast errors. The book closes with separate discussions of the “Box-Jenkins models” and Bayesian methods in forecasting. The appendices supply the usual statistical tables, a collection of actual time series data used in the examples in the text, and two computer programs for forecasting. The style of the book is direct in the sense that topics are presented in a well organized and sequenced manner. It is designed to be used at the level of an undergraduate course in forecasting