查詢結果分析
相關文獻
- Study for High-Order Models of Fuzzy Time Series
- Forecasting Enrollments with Fuzzy Time Series Using an Adaptive Partition
- 財務分析師,管理當局及統計模式預測準確度之比較研究
- A Refined Forecasting Model Based on Fuzzy Time Series and Particle Swarm Optimization
- 基於模糊時間序列和馬可夫模型預測入學人數與應用
- 臺灣房地產發展趨勢與價格預測模型之研究
- 語音壓縮標準G.723.1在C62X DSP上的設計
- 以臺灣在液化石油氣領域之優勢地位向華南輸出技術及成品之可行性研究
- 雙組織承諾之分類型態及影響因素探討
- 臺北市失業率之預測--模糊時間序列之應用
頁籤選單縮合
題 名 | Study for High-Order Models of Fuzzy Time Series=高階模糊時間序列模式之研究 |
---|---|
作 者 | 蔡朝治; | 書刊名 | 臺南師院學報 |
卷 期 | 35 2002.06[民91.06] |
頁 次 | 頁339-374 |
分類號 | 319 |
關鍵詞 | 模糊時間序列; 高階模式; 預測; 模糊關係; 語值; Fuzzy time series; High-order model; Forecasts; Fuzzy relationship; Linguistic value; |
語 文 | 英文(English) |
中文摘要 | 臺文提丑兩個定理和一個新的演算法。基於Song和Chissom的演算做預測,在理 論上無法釐清。所以,本文提出並證明兩個定理。基於此二定理建立一新的方法。此一新的 方法使時間序列模式獲得高階化。 為說明高階模式的效益,試將其演算結果與Chen的方法作比較。發現均方根誤差可由 Chen的方法之327 38降至該方法的59 75。又以台灣地區人口為例,均方根誤差可由Song和 Chissom的0 1190 降至該方法的013。 |
英文摘要 | This paper proposes two theorems and a new scheme for fuzzy series. In view of the work by Song and Chissom [4] 4, we find the model equation used is not theoretically clarified Hence, in this study, two theorems are proposed and proved to set as a basis. Then, a new scheme based on the theorems is proposed. This new scheme can be applied to models of the fuzzy time series and makes the models high-order. To illustrate the effect of this proposed new scheme, the forecasting enrollments are carried out. It is found that the root mean square error of the results can be improved from 327 38 for the Chen method to 59.75 for the current model, the model due to the injection of the new scheme into Chen model. Furthermore, forecasting population for Taiwan area is also implemented to reassure The results are compared with those of the other approaches. It is found that the root mean square error of the forecasts can be improved from 0.119 for the Song-Chissom method to 0.013 for the proposed model. |
本系統中英文摘要資訊取自各篇刊載內容。