面向高維數(shù)據(jù)的Takagi-Sugeno模糊系統(tǒng)建模新方法
doi: 10.11999/JEIT170792
國家自然科學(xué)基金(61300151),江蘇省自然科學(xué)基金(BK20160187, BK20161268),中央高?;究蒲袠I(yè)務(wù)費(fèi)專項(xiàng)項(xiàng)目(JUSRP11737)
A Novel Takagi-Sugeno Fuzzy Systems Modeling Method for High Dimensional Data
The National Natural Science Foundation of China (61300151), The Natural Science Foundation of Jiangsu Province (BK20160187, BK20161268), The Fundamental Research Funds for the Central Universities (JUSRP11737)
-
摘要: 對高維數(shù)據(jù)進(jìn)行建模是Takagi-Sugeno(T-S)模糊系統(tǒng)建模面臨的一個重大挑戰(zhàn)。為此,該文提出一種特征選擇與組稀疏編碼相結(jié)合的模糊系統(tǒng)建模新方法WOMP-GS-FIS。首先,運(yùn)用一種新型的加權(quán)正交匹配追蹤算法對原始樣本進(jìn)行特征選擇,在此基礎(chǔ)上提取出模糊規(guī)則前件并產(chǎn)生模糊系統(tǒng)字典;然后,基于組稀疏正則化構(gòu)造關(guān)于后件參數(shù)的組稀疏優(yōu)化問題,在優(yōu)化問題求解的同時得到重要的模糊規(guī)則。實(shí)驗(yàn)結(jié)果表明,在保證模型泛化性能的前提下,該方法不僅能對所獲得的模糊規(guī)則結(jié)構(gòu)進(jìn)行精簡還可以進(jìn)一步減少模糊規(guī)則數(shù),進(jìn)而解決高維數(shù)據(jù)環(huán)境下模糊規(guī)則可解釋性差的問題。
-
關(guān)鍵詞:
- T-S模糊系統(tǒng)建模 /
- 特征選擇 /
- 組稀疏編碼 /
- 精簡規(guī)則結(jié)構(gòu) /
- 模糊規(guī)則約減
Abstract: It is a great challenge to model Takagi-Sugeno(T-S) fuzzy systems on high dimensional data due to the problem of the curse of dimensionality. To this end, a novel T-S fuzzy system modeling method called WOMP-GS-FIS is proposed. The proposed method considers feature selection and group sparse coding simultaneously. Specifically, feature selection is performed by a novel Weighted Orthogonal Matching Pursuit (WOMP) method, based on which the fuzzy rule antecedent part is extracted and the dictionary of the fuzzy system is generated. Then, a group sparse optimization problem based on the group sparse regularization is formulated to obtain the optimal consequent parameters. In this way, the major fuzzy rules are selected by utilizing the group information that existing in the T-S fuzzy systems. The experimental results show that the proposed method can not only simplify the rule,s structure, but also reduce the number of fuzzy rules under the premise of good generalization performance, so as to solve the poor interpretation problem of fuzzy rules on high dimensional data effectively. -
程旸, 顧曉清, 蔣亦樟, 等. 具備視角協(xié)同學(xué)習(xí)能力的多視角TSK型模糊系統(tǒng)[J]. 電子與信息學(xué)報(bào), 2016, 38(8): 2054-2061. doi: 10.11999/JEIT151209. FERNNDEZ A, CARMONA C J, JESUS M J D, et al. A view on fuzzy systems for big data: Progress and opportunities[J]. International Journal of Computational Intelligence Systems, 2016, 9(s1): 69-80. doi: 10.1080/ 18756891.2016.1180820. CHENG Yang, GU Xiaoqing, JIANG Yizhang, et al. Multi- view TSK fuzzy system via collaborative learning[J]. Journal of Electronics Information Technology, 2016, 38(8): 2054-2061. doi: 10.11999/JEIT151209. LUO Minnan, SUN Fuchun, and LIU Huaping. Hierarchical structured sparse representation for T-S fuzzy systems identification[J]. IEEE Transactions on Fuzzy Systems, 2013, 21(6): 1032-1043. doi: 10.1109/TFUZZ.2013.2240690. JIANG Yizhang, DENG Zhaohong, CHUNG Fulai, et al. Recognition of epileptic EEG signals using a novel multiview TSK fuzzy system[J]. IEEE Transactions on Fuzzy Systems, 2017, 25(1): 3-20. doi: 10.1109/TFUZZ.2016.2637405. LUO Minnan, SUN Fuchun, and LIU Huaping. Joint block structure sparse representation for Multi-Input-Multi-Output (MIMO) T-S fuzzy system identification[J]. IEEE Transactions on Fuzzy Systems, 2014, 22(6): 1387-1400. doi: 10.1109/TFUZZ.2013.2292973. JUANG Chiafeng and HSIEH C D. TS-fuzzy system-based support vector regression[J]. Fuzzy Sets Systems, 2009, 160(17): 2486-2504. doi: 10.1016/j.fss.2008.11.022. JUANG Chiafeng and CHEN Guocyuan. A TS fuzzy system learned through a support vector machine in principal component space for real-time object detection[J]. IEEE Transactions on Industrial Electronics, 2012, 59(8): 3309-3320. doi: 10.1109/TIE.2011.2159949. 羅敏楠. T-S模糊推理系統(tǒng)的結(jié)構(gòu)稀疏編碼辨識理論與方法[D]. [博士論文], 清華大學(xué), 2014: 1-26. LUO Minnan. Theory and approches of T-S fuzzy inference systems identification with structure sparse coding[D]. [Ph.D. dissertation], Tsinghua University, 2014: 1-26. LUGHOFER E and KINDERMANN S. SparseFIS: data- driven learning of fuzzy systems with sparsity constraints[J]. IEEE Transactions on Fuzzy Systems, 2010, 18(2): 396-411. doi: 10.1109/TFUZZ.2010.2042960. SANA F, KATTERBAUER K, AL-NAFFOURI T Y, et al. Orthogonal matching pursuit for enhanced recovery of sparse geological structures with the ensemble kalman filter[J]. IEEE Journal of Selected Topics in Applied Earth Observations Remote Sensing, 2016, 9(4): 1710-1724. doi: 10.1109/JSTARS.2016.2518119. ZHOU Dengyong, BOUSQUET O, LAL T N, et al. Learning with local and global consistency[C]. Advances in Neural Information Processing Systems, Vancouver, Canada, 2004: 321-328. RODRGUEZ-FDEZ I, MUCIENTES M, and BUGARN A. Fruler: Fuzzy rule learning through evolution for regression [J]. Information Sciences, 2016, 354: 1-18. doi: 10.1016/j.ins. 2016.03.012. YUAN Ming and LIN Yi. Model selection and estimation in regression with grouped variables[J]. Journal of the Royal Statistical Society, 2006, 68(1): 49-67. doi: 10.1111/j.1467- 9868.2005.00532.x. ZHANG Caiya and XIANG Yanbiao. On the oracle property of adaptive group lasso in high-dimensional linear models[J]. Statistical Papers, 2016, 57(1): 249-265. doi: 10.1007/s00362- 015-0684-0. GRIGORIE L T and BOTEZ R M. Adaptive neuro-fuzzy inference system-based controllers for smart material actuator modelling[J]. Proceedings of the Institution of Mechanical Engineers Part G Journal of Aerospace Engineering, 2009, 223(6): 655-668. doi: 10.1243/09544100 JAERO522. NOROUZI J, YADOLLAHPOUR A, MIRBAGHERI S A, et al. Predicting renal failure progression in chronic kidney disease using integrated intelligent fuzzy expert system[J]. Computational Mathematical Methods in Medicine, 2016, 2016(3): 1-9. doi: 10.1155/2016/6080814. -
計(jì)量
- 文章訪問數(shù): 1736
- HTML全文瀏覽量: 211
- PDF下載量: 124
- 被引次數(shù): 0