A pseudo-Gaussian-based compensatory neural fuzzy system
Cheng‐Jian Lin, Wen-Hao Ho · 2004
In this paper, a new pseudo-Gaussian-based compensatory neural fuzzy system (PGCNFS) is proposed. The characteristic of compensatory neural fuzzy system is building exact fuzzy reasoning and converging quickly. Besides, the pseudo-Gaussian membership function can provide the compensatory neural fuzzy system which owns a higher flexibility and can approach the optimized result more accurately. An on-line learning algorithm is proposed to automatically construct the PGCNFS. It consists of structure learning and parameter learning that would create adaptive fuzzy logic rules. Experimental results show that the proposed algorithm converges quickly and the obtained fuzzy rules are more precise.