基於遺傳式橢圓形體分類演算法之不完整資料模糊分群

Fuzzy Clustering of Incomplete Data Based on Evolutionary Ellipsoid Classification Algorithm

黃成德(Cherng-Dir Huang)

指導教授 ： 姚立德

國立臺北科技大學/機電學院/機電整合研究所/碩士(2005年)

https://doi.org/10.6841/NTUT.2005.00104

摘要

傳統處理不完整資料的方法，最直接也最簡單的作法便是將不完整的部分忽略，但在分析某些資料量有限的原型資料集時，如生物的資料，卻往往會導致分群的效果不彰。因此，本論文利用遺傳式橢圓形體分類演算法(Evolutionary Ellipsoid Classification Algorithm, EECA)作為分析群聚問題之演算架構，並提出最短距離預測策略(Minimum Distance Strategy, MDS)，將不完整資料做有效的預測，並以UCI標準資料來驗證本文提出之方法。
最後將遺傳式橢圓形體分類演算法(EECA)應用於實際電力負載分類的問題上，並探討當某些負載特性曲線的特徵點遺失時，對演算法所造成的影響。

關鍵字

G-K演算法 ； 模糊群聚 ； 基因演算法 ； 不完整資料

並列摘要

Traditional methods of dealing with incomplete data are to ignore them. It is the simplest and most direct way. However, while analyzing a limited number of prototype data, it often results in ineffective grouping. In this thesis, the Evolutionary Ellipsoid Classification Algorithm (EECA) is proposed to find the best structure for the clustering problem, and the Minimum Distance Strategy (MDS), an effective prediction for the incomplete data. Moreover, several examples is also proposed as UCI standard databases are used to verify the effectiveness of the proposed approach.
Finally, we also employ the EECA in a real practice of load profile, and discover the influence caused by the algorithm, when some characteristics of load characteristic curves are lost.

並列關鍵字

G-K Algorithm ； Fuzzy Clustering ； Genetic Algorithm ； Incomplete Data

參考文獻

[1] J. K. Dixon, “Pattern recognition with partly missing data,” IEEE Trans. System, Man and Cybernetics, vol. 9, Oct. 1979, pp. 517-521.

[2] J. Kittler, “Classification of incomplete pattern vector using modified discriminant function,” IEEE Trans. Comput., vol. C-27, Apr. 1978, pp. 367-375.

[5] H. Timm C. Doring, and R. Kruse, “Different approaches to fuzzy clustering of incomplete datasets,” International Journal of Approximate, 2004, pp. 239-249.

[6] R. J. Hathaway and J. C. Bezdek, “Clustering incomplete relational data using the non-Euclidean relational fuzzy c-means algorithm,” Pattern Recognition Letter, 2002, pp. 151-160.

[7] B. Gabrys, “Pattern classification for incomplete data,” in Proc. Fourth International Conference on Knowledge-Based Intelligent Engineering Systems & Allied Technologies, Sept. 2000, pp. 454-457.