×

A novel generic Hebbian ordering-based fuzzy rule base reduction approach to Mamdani neuro-fuzzy system. (English) Zbl 1119.68152

Summary: There are two important issues in neuro-fuzzy modeling: (1) interpretability – the ability to describe the behavior of the system in an interpretable way – and (2) accuracy – the ability to approximate the outcome of the system accurately. As these two objectives usually exert contradictory requirements on the neuro-fuzzy model, certain compromise has to be undertaken. This letter proposes a novel rule reduction algorithm, namely, Hebb rule reduction, and an iterative tuning process to balance interpretability and accuracy. The Hebb rule reduction algorithm uses Hebbian ordering, which represents the degree of coverage of the samples by the rule, as an importance measure of each rule to merge the membership functions and hence reduces the number of the rules. Similar Membership Functions (MFs) are merged by a specified similarity measure in an order of Hebbian importance, and the resultant equivalent rules are deleted from the rule base. The rule with a higher Hebbian importance will be retained among a set of rules. The MFs are tuned through the least mean square algorithm to reduce the modeling error. The tuning of the MFs and the reduction of the rules proceed iteratively to achieve a balance between interpretability and accuracy. Three published data sets by Nakanishi [H. Nakanishi, I. B. Türksen, and M. Sugeno, Fuzzy Sets Syst. 57, 257–294 (1993; Zbl 1119.68454)], the Pat synthetic data set [S. K. Pal, S. Mitra, and P. Mitra (2003)], and the traffic flow density prediction data set are used as benchmarks to demonstrate the effectiveness of the proposed method. Good interpretability, as well as high modeling accuracy, are derivable simultaneously and are suitably benchmarked against other well-established neuro-fuzzy models.

MSC:

68T05 Learning and adaptive systems in artificial intelligence

Citations:

Zbl 1119.68454

Software:

RSPOP
PDFBibTeX XMLCite
Full Text: DOI

References:

[1] DOI: 10.1162/0899766052530857 · Zbl 1092.68637 · doi:10.1162/0899766052530857
[2] DOI: 10.1109/TSMCB.2003.812850 · doi:10.1109/TSMCB.2003.812850
[3] DOI: 10.1109/72.728363 · doi:10.1109/72.728363
[4] DOI: 10.1016/S0020-0255(01)00147-5 · Zbl 0996.68158 · doi:10.1016/S0020-0255(01)00147-5
[5] DOI: 10.1109/TFUZZ.2004.839670 · Zbl 05452487 · doi:10.1109/TFUZZ.2004.839670
[6] DOI: 10.1109/21.370193 · doi:10.1109/21.370193
[7] DOI: 10.1109/TNN.2003.820557 · doi:10.1109/TNN.2003.820557
[8] DOI: 10.1109/91.660804 · doi:10.1109/91.660804
[9] DOI: 10.1109/91.855921 · doi:10.1109/91.855921
[10] DOI: 10.1016/S0165-0114(96)00280-1 · Zbl 05468312 · doi:10.1016/S0165-0114(96)00280-1
[11] DOI: 10.1109/3477.931534 · doi:10.1109/3477.931534
[12] DOI: 10.1109/21.256541 · doi:10.1109/21.256541
[13] DOI: 10.1109/91.660805 · doi:10.1109/91.660805
[14] DOI: 10.1109/3477.969494 · doi:10.1109/3477.969494
[15] DOI: 10.1109/91.995117 · doi:10.1109/91.995117
[16] DOI: 10.1007/BF00337288 · Zbl 0466.92002 · doi:10.1007/BF00337288
[17] DOI: 10.1016/S0165-0114(00)00112-3 · Zbl 1023.62007 · doi:10.1016/S0165-0114(00)00112-3
[18] DOI: 10.1016/S0020-7373(75)80002-2 · Zbl 0301.68076 · doi:10.1016/S0020-7373(75)80002-2
[19] DOI: 10.1109/72.846746 · doi:10.1109/72.846746
[20] DOI: 10.1016/S0925-2312(00)00335-0 · Zbl 1003.68636 · doi:10.1016/S0925-2312(00)00335-0
[21] DOI: 10.1016/0165-0114(93)90024-C · Zbl 1119.68454 · doi:10.1016/0165-0114(93)90024-C
[22] DOI: 10.1109/TKDE.2003.1161579 · Zbl 05108607 · doi:10.1109/TKDE.2003.1161579
[23] DOI: 10.1109/3477.809038 · doi:10.1109/3477.809038
[24] DOI: 10.1109/5326.941843 · doi:10.1109/5326.941843
[25] DOI: 10.1109/3477.678632 · doi:10.1109/3477.678632
[26] DOI: 10.1109/3477.891144 · doi:10.1109/3477.891144
[27] DOI: 10.1016/0165-0114(88)90113-3 · Zbl 0652.93010 · doi:10.1016/0165-0114(88)90113-3
[28] DOI: 10.1109/TFUZZ.1993.390281 · doi:10.1109/TFUZZ.1993.390281
[29] DOI: 10.1109/TSMC.1985.6313399 · Zbl 0576.93021 · doi:10.1109/TSMC.1985.6313399
[30] DOI: 10.1109/TNN.2002.1031940 · doi:10.1109/TNN.2002.1031940
[31] DOI: 10.1109/3477.740162 · doi:10.1109/3477.740162
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.