Loading...
|
Please use this identifier to cite or link to this item:
https://nccur.lib.nccu.edu.tw/handle/140.119/67472
|
| Title: | 以狀態空間模型分析常見的經濟資料
The Analysis of Economic Data by State-space Model |
| Authors: | 莊雅婷 |
| Contributors: | 翁久幸
莊雅婷 |
| Keywords: | 狀態空間模型
卡爾曼濾波器
最大概似估計法
狀態變數 |
| Date: | 2013 |
| Issue Date: | 2014-07-14 11:29:52 (UTC+8) |
| Abstract: | 在統計中有許多估計參數的方法，像是點估計中的動差法、最大概似估計法、最小平方法，和區間估計等等。在動態的時間序列裡常會使用到狀態空間模型(State–space model)來處理動態時間序列模型中未觀測到的狀態變數。在計量經濟中常常會包含許多未觀測到的狀態變數，例如：事前實質利率、永久所得等等，因此狀態空間模型在計量經濟學中有很廣泛的運用。
本論文以狀態空間模型(State－space model)表示模型，再以卡爾曼濾波器(Kalman filter)為輔助分析工具，藉以分析三筆常見的經濟資料：美國實質國內生產毛額(U.S. Real GDP)、西德州中級原油價格(West Texas Intermediate Crude Oil)，以及國際黃金價格，想從分析的資料中得知觀測值的變異主要源自模型中的隨機趨勢項或是平穩循環項。
得到的結論為：在美國實質國內生產毛額(U.S. Real GDP)資料裡發現資料研究期間的長短，所得到的結論會有所不同，在短期時間的資料裡，觀測值的變異主要源自平穩循環項，而在長期時間的資料裡則相反；在西德州中級原油價格(West Texas Intermediate Crude Oil)，以及國際黃金價格的資料分析中，得知觀測值的變異主要來自隨機趨勢項。
There are many methods to estimate parameter in statistic, such as the method of moment, maximum likelihood estimation (MLE), least squares estimates (LSE), and interval estimation etc. State-space models always deal with the dynamic time series models that have unobserved state variables. Econometrics often involves many unobserved state variables, for instance, the ex ante real rate of interest, permanent income etc. State-space models have a wide range of applications in econometrics.
This study analysis three common economic data, U.S real gross domestic product, West Texas intermediate crude oil price, and gold price, by using the state-space model to represent the time series model, and the Kalman filter, the basic tool to deal with the state-space model.
In U.S real GDP data, the different research period may affect the conclusions. In the short period, the variation of observed data is mainly from the stationary cyclical component. On the contrary, in the long period, the variation of observed data is mainly from the stochastic trend term. In WTI oil price and gold price data, the variation of observed data is mainly from stochastic trend component. |
| Reference: | 1. CLARK, Peter K. The cyclical component of US economic activity. The Quarterly Journal of Economics, 1987, 797-814.
2. HARVEY, A. C. Time Series Models. 1981.
3. KALMAN, Rudolph Emil. A new approach to linear filtering and prediction problems. Journal of Fluids Engineering, 1960, 82.1: 35-45.
4. KIM, Chang-Jin; NELSON, Charles R. State-space models with regime switching. Cambridge, Mass, 1999.
5. NELSON, Charles R.; PLOSSER, Charles R. Trends and random walks in macroeconmic time series: some evidence and implications. Journal of monetary economics, 1982, 10.2: 139-162.
6. WELCH, Greg; BISHOP, Gary. An introduction to the Kalman filter. 1995. |
| Description: | 碩士
國立政治大學
統計研究所
101354023
102 |
| Source URI: | http://thesis.lib.nccu.edu.tw/record/#G0101354023 |
| Data Type: | thesis |
| Appears in Collections: | [統計學系] 學位論文
|
Files in This Item:
| File |
Size | Format | |
|---|
| 402301.pdf | 549Kb | Adobe PDF2 | 394 | View/Open |
|
All items in 政大典藏 are protected by copyright, with all rights reserved.
|
