查詢結果分析
來源資料
相關文獻
- 月流量序列補遺之研究
- 教育測驗資料Missing Data的介紹及處理方法簡介
- 我國航空運輸業就業人口供需之研究
- The Response of Stock Prices to Permanent and Transitory Shocks to Accounting Earnings
- 自我迴歸移動平均整合模式之參數估計--介紹BHHH法及其實證
- Time Series Forecasting of Kaohsiung Unemployment Rate Using Neural Network Model
- Optimal Control of Eutrophication for a Reservoir: The Linear Quadratic Regulation Approach
- 線性結構關係的時間序列模式
- 時間序列模式分析實例
- 時間序列模式分析實例
頁籤選單縮合
題名 | 月流量序列補遺之研究 |
---|---|
作者姓名(中文) | 虞國興; 魏柏村; | 書刊名 | 農業工程學報 |
卷期 | 41:3 1995.09[民84.09] |
頁次 | 頁14-23 |
分類號 | 351.8 |
關鍵詞 | 遺失值; 時間序列模式; Missing data; Time series model; |
語文 | 中文(Chinese) |
中文摘要 | 水文資料常因儀器、人為、天災等因素發生遺失或錯誤之現象,致使水文分析上發生不少困擾,故遺失資料之補遺為一十分重要之研究課題,本研究以定常性合成資料驗證Ljung (1989)、Pourahmadi (1989)及Abrahan (1981)所提方法之精確度,並探討其於實測資料之適用性。 研究結果顯示,Ljung、Abrahan及Pourahmadi三種方法之精確度大致相同,然Abrahan方法偶有發散現象,配合Yu及Lin (1991)之部份自迴歸模式應用於臺灣實測月流量資料之補遺,以本研究所提之修正Abrahan方法精確度最高。另,研究中亦探討具遺失值時序模式判定之正確性,研究結果顯示,模式之選取準則以BIC較佳。 |
英文摘要 | Due to the instruments malfunction, human factor or disaster, hydrological data are often missing. It causes lots of problem in the analysis. Therefore, the estimation of missing data plays an important role in the hydrological analysis. In the present study, stationary synthetic data are used to investigate the accuracy of estimating the missing proposed by Ljung (1989), Pourahmadi (1989) and Abraham (1981). Meanwhile, its aptness for analysis the real data is also investigated. The results indicate that Ljung, Abraham and Pourahmadi are with the same accuracy. However, there is divergent suitation for Abraham method. For the estimation of he missing data of monthly streamflow, the SAR model proposed by Yu and Lin (1991) is used for modeling. The modified Abraham method proposed in the present study gives the highest accuracy for the monthly streamflow in Taiwan. Besides, the accuracies of model selection criteria are also investigated for the time series with some missing data. The results indicate the BIC serves the better performance of model selection. |
本系統之摘要資訊系依該期刊論文摘要之資訊為主。