This research introduces a new data-driven methodology for fault detection and isolation in dynamic systems, integrating fuzzy/Bayesian change point detection and decision trees induced by genetic programming for pattern classification. Tracking changes in sensor signals enables the detection of faults, and using decision trees generated by genetic programming allows for accurate categorization into specific fault classes. Change point detection utilizes a combination of fuzzy set theory and the Metropolis–Hastings algorithm. The primary contribution of the study lies in the development of a distinctive classification system, which results in a comprehensive and highly effective approach to fault detection and isolation. Validation is carried out using the Tennessee Eastman benchmark process as an experimental framework, ensuring a rigorous evaluation of the efficacy of the proposed methodology.