Using a memristor crossbar structure to implement a novel adaptive real-time fuzzy modeling algorithm

被引:16
作者
Afrakoti, Iman Esmaili Paeen [1 ]
Shouraki, Saeed Bagheri [2 ]
Bayat, Farnood Merrikh [2 ]
Gholami, Mohammad [1 ]
机构
[1] Univ Mazandaran, Fac Engn & Technol, Babol Sar, Iran
[2] Sharif Univ Technol, Res Grp Brain Simulat & Cognit Sci, Artificial Creatures Lab, Sch Elect Engn, Azadi Ave, Tehran, Iran
关键词
Active learning method; Fuzzy inference; Memristor crossbar; Optimization-free; Pattern classification; IDS METHOD; SYSTEMS;
D O I
10.1016/j.fss.2016.02.016
中图分类号
TP301 [理论、方法];
学科分类号
081202 ;
摘要
Fuzzy techniques can be used for accurate and high-speed modeling as well as for the control of complex systems, but various challenging problems are usually encountered during their actual implementation. For example, the variable parameters need to be optimized iteratively during the training phase, where this process is inspired by crisp domain algorithms. However, in recent years, memristor-based structures have emerged as another promising method for implementing neural network structures and fuzzy algorithms. In this study, we propose a novel adaptive and real-time fuzzy modeling algorithm, which employs the active learning method concept to mimic the functionality of the brain's right hemisphere. The proposed method processes fuzzy numbers such that the system retains its sensitivity to individual training data points, which expands the knowledge tree to allow plasticity, as well as using a specific defuzzification technique that guarantees stability. Another advantage of our new processing engine is that the nature of processing within the fuzzy system is consistent with that observed in memristive devices. Thus, we demonstrate that the proposed fuzzy architecture can be implemented easily and efficiently using existing crossbar structures. We verified the effectiveness of the proposed algorithm in modeling and pattern recognition tasks based on computer simulations. (C) 2016 Elsevier B.V. All rights reserved.
引用
收藏
页码:115 / 128
页数:14
相关论文
共 27 条
[1]   An Optimal Hardware Implementation for Active Learning Method Based on Memristor Crossbar Structures [J].
Afrakoti, Iman Esmaili Paeen ;
Shouraki, Saeed Bagheri ;
Haghighat, Bahar .
IEEE SYSTEMS JOURNAL, 2014, 8 (04) :1190-1199
[2]  
[Anonymous], 2 INT S INT INF TECH
[3]   MEMRISTOR - MISSING CIRCUIT ELEMENT [J].
CHUA, LO .
IEEE TRANSACTIONS ON CIRCUIT THEORY, 1971, CT18 (05) :507-+
[4]  
de Garis H., 1994, J SOC INSTR CONTROL, V33, P2
[5]   SUPERIOR GENERALIZATION CAPABILITY OF HARDWARE-LEARNING ALGORITHM DEVELOPED FOR SELF-LEARNING NEURON-MOS NEURAL NETWORKS [J].
KONDO, S ;
SHIBATA, T ;
OHMI, T .
JAPANESE JOURNAL OF APPLIED PHYSICS PART 1-REGULAR PAPERS SHORT NOTES & REVIEW PAPERS, 1995, 34 (2B) :1066-1069
[6]  
Kuekes P., 2008, MEMR MEMR SYST S NOV
[7]  
Lang KJ., 1988, Proc. of 1988 Connectionist Models Summer School, ppp 52
[8]  
Linares-Barranco B., 2009, Memristance can explain SpikeTime-Dependent-Plasticity in Neural Synapses
[9]   Memristive Neuro-Fuzzy System [J].
Merrikh-Bayat, Farnood ;
Shouraki, Saeed Bagheri .
IEEE TRANSACTIONS ON CYBERNETICS, 2013, 43 (01) :269-285
[10]   Memristor Crossbar-Based Hardware Implementation of the IDS Method [J].
Merrikh-Bayat, Farnood ;
Shouraki, Saeed Bagheri ;
Rohani, Ali .
IEEE TRANSACTIONS ON FUZZY SYSTEMS, 2011, 19 (06) :1083-1096