| Automatic Technology |
|
|
|
|
| Incremental zero-order TSK fuzzy classifier and its robust version |
| LI Tao, WANG Shi-tong |
| School of Digital Media, Jiangnan University, Wuxi 214122, China |
|
|
|
Abstract An incremental zero-order TSK fuzzy classifier called TSK-IFCIRLS0 was proposed based on iteratively reweighted least squares optimization algorithm in order to circumvent the drawback that traditional non-fuzzy classifiers had no any interpretability and that fuzzy classifiers could not always be feasible for many datasets with satisfied classification performance. The incremental fuzzy clustering algorithm IFCM(c+p) for large-scale datasets was used to quickly train antecedent parameters of fuzzy rules by clustering and using Gauss function to map the clustering results into fuzzy subspace. The iteratively reweighted least squares optimization algorithm was used to learn consequent parameters of fuzzy rules. The robust version called TSK-IFCPHub0 was developed based on pseudo-Huber loss function with the purpose of improving anti-noise ability of TSK-IFCIRLS0. The proposed fuzzy classifiers were experimentally compared with conventional fuzzy classifier FCPM-IRLS, RBF neural network and ANFIS. Results indicated the power of the proposed fuzzy classifiers on interpretability, classification performance and scalability. The strong robust capability of TSK-IFCPHub0 was verified by the experimental results.
|
|
Received: 31 August 2016
Published: 27 September 2017
|
|
|
增量式0阶TSK模糊分类器及鲁棒改进
为了克服传统的分类器难以在具有令人满意的分类性能、快速的学习效率的同时兼顾高可解释性之不足，提出增量式0阶模糊分类器TSK-IFCIRLS0.该分类器通过使用增量式模糊聚类算法IFCM（c+p）对训练样本进行聚类，使用高斯隶属度函数将聚类结果映射到模糊子空间，使用迭代重加权最小二乘优化算法IRLS对模糊规则的后件参数进行学习.通过提出基于伪Huber函数的代价函数，它的鲁棒性改进版本TSK-IFCPHub0被提出来以提高分类器的抗噪能力.仿真实验表明，与FCPM-IRLS、RBF、ANFIS分类器相比，提出的2种模糊分类器均具有良好的分类性能及数据规模的可扩展性，TSK-IFCPHub0具有良好的鲁棒性.
|
|
| [1] HEARST M A. Support vector machines[J]. IEEE Intelligent Systems, 1998, 13(4):18-28.
[2] HUANG G B, MAO K Z, SIEW C K, et al. Fast modular network implementation for support vector machines[J]. IEEE Transactions on Neural Networks, 2005, 16(6):1651-1663.
[3] SUGUMARAN V, RAMACHANDRAN K I. Effect of number of features on classification of roller bearing faults using SVM and PSVM[J]. Expert Systems with Applications, 2011, 38(4):4088-4096.
[4] HUANG G B, ZHU Q Y, SIEW C K. Extreme learning machine:theory and applications[J]. Neurocomputing, 2006, 70(1):489-501.
[5] JAFARZADEH S, FADALI M S, SONBOL A H. Stability analysis and control of discrete type-1 and type-2 TSK fuzzy systems:Part Ⅱ. control design[J]. IEEE Transactions on Fuzzy Systems, 2011, 19(6):1001-1013.
[6] FADALI M S, JAFARZADEH S. TSK observers for discrete type-1 and type-2 fuzzy systems[J]. IEEE Transactions on Fuzzy Systems, 2014, 22(2):451-458.
[7] 蔡前凤,郝志峰,杨晓伟.基于核映射的高阶Takagi-Sugeno模糊模型[J].控制理论与应用, 2011,28(5):681-687. CAI Qian-feng, HAO Zhi-feng, YANG Xiao-wei. Higher-order Takagi-Sugeno fuzzy model based on kernel mapping[J]. Control Theory and Applications, 2011, 28(5):681-687.
[8] 杨昌健,邓赵红,王士同,等.基于0阶TSK型迁移模糊系统的EEG信号自适应识别[J].计算机应用研究,2015, 32(8):2276-2280. YANG Chang-jian, DENG Zhao-hong, WANG Shi-tong, et al. Adaptive recognition of epileptic EEG signals based on 0-order TSK type transfer fuzzy system[J]. Application Research of Computers, 2015, 32(8):2276-2280.
[9] 邓赵红,张江滨,王士同,等.基于模糊子空间聚类的0阶L2型TSK模糊系统[J].电子与信息学报, 2015,37(9):2082-2088. DENG Zhao-hong, ZHANG Jiang-bin, WANG Shi-tong, et al. Fuzzy subspace clustering based zero-order L2-norm TSK fuzzy system[J]. Journal of Electronics and Information Technology, 2015, 37(9):2082-2088.
[10] LESKI J M. Fuzzy (c+p)-means clustering and its application to a fuzzy rule-based classifier:towards good generalization and good interpretability[J]. IEEE Transactions on Fuzzy Systems, 2015, 23(4):802-812.
[11] STARCZEWSKI J T, NOWICKI R K, NOWAK B A. Genetic fuzzy classifier with fuzzy rough sets for imprecise data[C]//IEEE International Conference on Fuzzy Systems. Beijing:IEEE, 2014:1382-1389.
[12] 刘淑英.混合神经模糊分类器的实现[J].计算机技术与发展, 2013, 12(12):113-115. LIU Shu-ying. Hybrid neural fuzzy classifier to achieve[J]. Computer Technology and Development, 2013,12(12):113-115.
[13] YANG X, ZHANG G, LU J, et al. A kernel fuzzyc-means clustering-based fuzzy support vector machine algorithm for classification problems with outliers or noises[J]. IEEE Transactions on Fuzzy Systems, 2014, 19(1):105-115.
[14] 李滔,王士同.适合大规模数据集的增量式模糊聚类算法IFCM(c+p)[J].智能系统学报, 2016, 11(2):227-233. LI Tao, WANG Shi-tong. Incremental fuzzy (c+p)-means clustering for large data[J]. CAAI Transactions on Intelligent Systems, 2016, 11(2):227-233.
[15] HARTLEY R, ZISSERMAN A. Multiple view geometry in computer vision[J]. Cambridge University Press, 2006, 30(9-10):1865-1872.
[16] 王立新.模糊系统与模糊控制教程[M].北京:清华大学出版社, 2003:81-91.
[17] 李航.统计学习方法[M].北京:清华大学出版社, 2012:210-224.
[18] 朱嘉钢,王士同.Huber-SVR中参数μ与输入噪声间关系的研究[J].复旦学报:自然科学版,2004,43(5):793-796. ZHU Jia-gang, WANG Shi-tong. Research on the dependency between μ and the input noise in Huber-support vector regression[J].Journal of Fudan University:Natural Science, 2004, 43(5):793-796.
[19] SUN H, WANG S, JIANG Q. FCM-based modelselection algorithms for determining the number of clusters[J]. Pattern recognition, 2004, 37(10):2027-2037.
[20] YU J, CHENG Q, HUANG H. Analysis of the weighting exponent in the FCM[J]. IEEE Transactions on Systems, Man, and Cybernetics, Part B:Cybernetics, 2004, 34(1):634-639.
[21] JANG J S R. ANFIS:adaptive-network-based fuzzy inference system[J]. IEEE Transactions on Systems Man and Cybernetics, 1993, 23(3):665-685.
[22] RATSCH G, ONODA T, MULLER K R. Soft margins for AdaBoost[J]. Machine Learning, 2001,42(3):287-320. |
|
Viewed |
|
|
|
Full text
|
|
|
|
|
Abstract
|
|
|
|
|
Cited |
|
|
|
|
| |
Shared |
|
|
|
|
| |
Discussed |
|
|
|
|
