Order estimation of high dimensional time series
Much research has focused on the problem of estimating the order of vector autoregressive (VAR) model and multivariate moving average (VMA) model. The most proposed solutions for this problem include Bayesian Information Criterion (BIC) and limiting spectral distribution of sample autocovariance mat...
Saved in:
| 主要作者: | |
|---|---|
| 其他作者: | |
| 格式: | Final Year Project |
| 語言: | English |
| 出版: |
Nanyang Technological University
2022
|
| 主題: | |
| 在線閱讀: | https://hdl.handle.net/10356/156931 |
| 標簽: |
添加標簽
沒有標簽, 成為第一個標記此記錄!
|
| 機構: | Nanyang Technological University |
| 語言: | English |
| 總結: | Much research has focused on the problem of estimating the order of vector autoregressive (VAR) model and multivariate moving average (VMA) model. The most proposed solutions for this problem include Bayesian Information Criterion (BIC) and limiting spectral distribution of sample autocovariance matrix. In this paper, two new approaches for order determination of VAR and VMA model are proposed. Maximum or sum of eigenvalues of sample autocovariance matrix is found to be able to select the order of VAR or VMA model. Another approach with the use of information criterion is also developed to select the order automatically. Time series data has been generated to examine the performance of two methods in existing work and two approaches proposed by ourselves. |
|---|