Abstract

In this paper, the segmented and breakpoint regression models were estimated using the Ordinary Least Squares (OLS) method and the Adaptive Neuro fuzzy Inference model (ANFIS) under a dataset of (200) patients. The dependent variable (y) cardiovascular complications, six explanatory variables (age, body mass index, blood glucose level, systolic blood pressure, cholesterol level, and C-reactive protein). The Chow test was first performed to identify regression points, revealing two points (110 and 160). The ANFIS model proved superior to the Ordinary Least Squares (OLS) method in estimating the risk of complications. Furthermore, these systems provided accurate estimates that accurately reflected the studied phenomenon, unlike the OLS method, which yielded inaccurate and unreliable estimates. The ANFIS estimates also proved that all research variables were significant and had an impact on the occurrence of complications resulting from cardiovascular disease.