Performance analysis of adaptive neuro-fuzzy inference-based relying on modeling electric modulus feature of 3-formylchromone

Abstract

The performance of ANFIS (Adaptive Neuro-Fuzzy Inference System) for modelling and foretell 3-formylchromone's electric modulus conduct was explored. Based on experimental data of the M/, M//; real and imaginary components of the M*, complex electric modulus in temperatures portions (298–388 K), two mathematical models (empirical expressions) were developed. The data divided into training and testing two groups by ratio 80% and 20% of dataset, respectively.In ANFIS, neural networks are combined with fuzzy logic to advance prediction ability. A hybrid learning technique was employed to construct ANFIS models. The training as well as testing findings was in accordance with the experimental results. The outcomes revealed that the correlation determination coefficient was the highest and the root mean square error was the lowest. The findings support the ANFIS model's capacity to simulate and anticipate the electric modulus conduct in 3-formylchromone.