Assessing the performance of artificial intelligence regression models for predicting the fatigue of a bio-based composite

The demand for natural fiber-reinforced composites in engineering design has increased, offering advantages like low density and environmental benefits. However, their long-term behavior, particularly under cyclic loading, remains limited. This study aims to predict the fatigue behavior of composites reinforced with natural fiber using several predictive models of regression trained on a database of studies. This model reduces experimental testing and improves prediction quality by avoiding computationally intensive data points.