Studies on Parameter Identification Using Fuzzy Random Data
Grant-in-Aid for Scientific Research (C)
|Allocation Type||Single-year Grants|
|Research Institution||OTEMON GAKUIN UNIVERSITY|
FUKUDA Tokuo OTEMON GAKUIN UNIVERSITY,Faculty of Management, Professor, 経営学部, 教授 (60201745)
|Project Period (FY)
1996 – 1998
Completed(Fiscal Year 1998)
|Budget Amount *help
¥1,400,000 (Direct Cost : ¥1,400,000)
Fiscal Year 1998 : ¥500,000 (Direct Cost : ¥500,000)
Fiscal Year 1997 : ¥500,000 (Direct Cost : ¥500,000)
Fiscal Year 1996 : ¥400,000 (Direct Cost : ¥400,000)
|Keywords||Fuzzy Random Vectors / Set Representations of Fuzzy Sets / Multi-valued Logic / Theory of Correspondences / Statistical Moments / 曖昧不規則データ / 曖昧推定量 / 大数の法則 / Hausdorff距離 / 多値理論 / 対応の理論|
In this research, a class of fuzzy random vectors(FRVCs) and their statistical moments have been introduced from the consistent viewpoint of vague perceptions of random phenomena, and some properties of FRVCs and their statistical moments up to second ones have been examined theoretically.
The features of proposed FRVCs are as follows :
1. Set representations of fuzzy sets are thoroughly adopted because of ilie feasibility for describing operations between FRVCs.
2. By adopting the family of compact convex correspondences as the set representation of a FRVCS, we have been able to introduce the reasonable definition of expectations of FRVCs, which is basically compatible with (scalar) fuzzy random variables proposed by Kwakernaak and is never obtained from their simple extensions.
3. The proposed concept of FRVCs as vague perceptions of random vectors may be, in some sense, a bridge between two major types of flizzy mndoin variables.
Furthermore, the following items should be investigated in the future :
1. Definitions of higher statistical moments of FRVCs and their estimators.
2. Statistical properties of FRVCS under more moderate conditions
3. Applications of the proposed approach to fuzzy random data obtained by vague perceptions of output processes of time series described by AR, MA and ARMA models and so on.
4. Identification of system parameters of time series models as mentioned above bv using fuzzy random data.
Research Output (21results)