| λ—模糊测度辨识方法研究 |
| |
| 引用本文： | 孙怀江,杨静宇.λ—模糊测度辨识方法研究[J].南京理工大学学报(自然科学版),2001,25(1):44-47. |
| |
| 作者姓名： | 孙怀江 杨静宇 |
| |
| 作者单位： | 南京理工大学计算机科学与技术系, |
| |
| 基金项目： | 国防科技预研行业基金项目 |
| |
| 摘 要： | 研究了当专家给出的模糊测度具有不确定性时λ模糊测度的辨识问题，不确定数据分别用区间和模糊数来表达，针对不同情况提出了4种误差函数，误差函数的最小值对应于最优的辨识结果，而且当不存在的不确定性时，所得结果与目前已有方法相同。数值实验结果表明，该文提出的辨识方法是有效的。
|
| 关 键 词： | 人工智能 模式识别 λ－模糊测度 误差函数 梯度算法 辨识方法 |
| 修稿时间： | 2000年3月16日 |
A Study on Identification
Methods of |
| |
| SunHuaijiang,YangJingyu.A Study on Identification
Methods of[J].Journal of Nanjing University of Science and Technology(Nature Science),2001,25(1):44-47. |
| |
| Authors: | SunHuaijiang YangJingyu |
| |
| Abstract: | Fuzzy measures given by human experts usually contain uncertainties, and do not satisfy required mathematical properties. The problem of identifying lambda fuzzy measures is studied when the uncertainties are expressed by intervals and fuzzy numbers. Four error functions are proposed. The minima of error functions correspond to optimal identified results. When there is not any uncertainty, obtained results are equal to those from existing methods. Numerical experimental results indicate that the proposed identification algorithms are efficient. |
| |
| Keywords: | artificial intelligence pattern recognition measure |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
