AN EMPIRICAL EVALUATION OF THE PERFORMANCE OF TRAMO / SEATS ON SIMULATED SERIES
AN EMPIRICAL EVALUATION OF THE PERFORMANCE OF TRAMO / SEATS ON SIMULATED SERIES
复制标题
TRAMO / SEATS 在模拟系列上性能的实证评估
DOI：
--
复制
发表时间：
2000
期刊：
影响因子：
--
通讯作者：
D. Findley
中科院分区：
文献类型：
--
作者：
C. C. Hood;James D. Ashley;D. Findley
TRAMO（具有ARIMA噪声、缺失观测值和异常值的时间序列回归）和SEATS（ARIMA时间序列中的信号提取）是由维克托·戈麦斯和奥古斯丁·马拉瓦尔开发的相关程序，用于使用基于ARIMA模型的信号提取技术对时间序列进行季节性调整。我们将评估TRAMO/SEATS在一些模拟经济时间序列上的性能，包括具有较大不规则成分的序列、具有复杂趋势的序列以及短序列。
统计局正在转向一种基于机构调查的新分类系统。原标准产业分类（SIC）正被北美产业分类系统（NAICS）所取代。当这种情况发生时，人口普查局的一些序列有可能仅需用几年的数据进行调整。所以我们将研究仅用四年数据对我们的模拟序列进行调整。
我们评估了TRAMO和SEATS的季节性调整性能，并与X - 12 - ARIMA的调整进行了比较。X - 12 - ARIMA是人口普查局在X - 11系列季节性调整程序中的最新程序。我们研究了SEATS的三个不同版本，包括一个可以使用SEATS算法的X12 - ARIMA版本，以下称为X - 12 - ARIMA - SEATS。X - 12 - ARIMA - SEATS使我们能够为两个程序计算类似的诊断信息，这样我们就可以比较基于模型的调整和X - 12方法之间的调整。
TRAMO (Time series Regression with ARIMA noise, Missing observations, and Outliers) and SEATS (Signal Extraction in ARIMA Time Series) are linked programs developed by Victor Gomez and Agustin Maravall to seasonally adjust time series using ARIMA model-based signal extraction techniques. We will evaluate the performance of TRAMO/SEATS on some simulated economic time series, including series with a large irregular component, series with complex trends, and short series. The Bureau is moving to a new classification system for establishment based surveys. The former Standard Industrial Codes (SIC) are being replaced by the North American Industrial Classification System (NAICS). When this happens, there is a possibility that some series at the Census Bureau will need to be adjusted with only a few years of data. So we will look at adjusting our simulated series with only four years of data. We evaluated the performance of seasonal adjustments from TRAMO and SEATS as compared to adjustments from X-12-ARIMA. X-12-ARIMA is the Census Bureau's latest program in the X-11 line of seasonal adjustment programs. We looked at three different versions of SEATS, including a version of X12-ARIMA that has access to the SEATS algorithm hereafter called X-12-ARIMA-SEATS. X-12-ARIMASEATS allows us to compute similar diagnostics for both programs so we can compare adjustments between modelbased adjustment and the X-12 method.