Combining neural fuzzy inference system and machine learning algorithms to improve fabric prediction models using technical parameters.

Abstract

The textile industry plays a critical role in daily life, but faces challenges such as high levels of waste and environmental impact. This study explores the integration of Artificial Intelligence (AI) and machine learning techniques to address these issues, focusing on predictive modeling of fabric properties. Specifically, the study develops and validates an Adaptive Neuro-Fuzzy Inference System (ANFIS) model using real and synthetic data generated by Conditional Generative Adversarial Networks (CTGAN). CTGAN augments the dataset while preserving privacy, and the Discrete Wavelet Transform (DWT) aids in outlier detection to ensure data integrity. The research demonstrates the efficacy of AI-driven approaches in improving fabric property prediction accuracy, reducing waste, and enhancing production efficiency. The findings encourage broader adoption of AI technologies in the textile industry for sustainable and efficient manufacturing processes. Keywords: Textile industry, Artificial Intelligence, ANFIS, CTGAN, predictive modeling, waste management