A Novel Fuzzy System With Adaptive Neurons for Earthquake Modeling

Abstract

Data driven fuzzy neural networks have some disadvantages, such as high dimensions and complex learning process. Also, the obtained models are difficult to interpret. In this paper, we propose a novel simple fuzzy system, which uses fuzzy adaptive neurons. This novel model takes the advantages of the interpretability of the fuzzy system and good approximation ability of the neural networks. We propose a simple learning algorithm for the novel fuzzy system. The stability analysis is given. We successfully apply this fuzzy model for the earthquake modeling. Comparisons with the popular fuzzy neural model are proposed.

A New Spike Membership Function for the Recognition and Processing of Spatiotemporal Spike Patterns: Syllable-Based Speech Recognition Application

“…The temporal-based approach effectively improves the accuracy of visual and acoustic pattern recognition tasks. Activation functions or membership functions with different forms, such as those by Sigmoid, Gaussian, Parabola, and Sigmoid-Gaussian, among others, have been developed in Zhao and Bose, Basterretxea et al, Xie et al, and Ramirez-Mendoza et al [26][27][28][29][30][31][32].…”

“…Different membership functions and approximations of activation functions have been proposed for various applications, such as seismic modeling of accelerograms and control, among others [26][27][28][29][30][31][32]. Following is the development of a spike membership function for unipolar inputs z in j , shown in Figure 1.…”

“…Spike patterns recognition of the SI syllable in a melody is performed by obtaining the maximum of the STFT magnitudes and the Augmented FAN-STEPAF-SPKAF method [29][30][31][32][33][34][35].…”

“…According to the study presented in [31] for the stability analysis of adaptive neurons, it is considered that a nonlinear system in discrete time can be represented as…”

“…Remark 2. FAN-SPKAF uses the same learning algorithm for unipolar fuzzy systems, ∈ [0,1], developed for FAN-SAF in [16,31,33]. Because the spikes are bipolar, to normalize them to the unipolar interval, a max operator and a zero reference or a comparator with square output pulses ∈ [0, 1] could be used.…”