RESUMEN

Decision-makers have consistently developed a range of classification models, each possessing unique features within the domain of intelligent models. These endeavors are all directed toward achieving the highest levels of accuracy. In recent developments, two notable methodologies-reliable modeling and jumping modeling approaches-offer specific advantages in formulating cost functions and have been recognized for their role in enhancing classifier accuracy. Specifically, the jumping methodology is based on aligning the learning process with the discrete nature of the classification goal, while the reliable methodology integrates the reliability factor into the learning paradigm. However, their innovative combination, leveraging both accuracy and reliability factors in guiding learning processes, leads to the creation of a high-performing classifier. This addresses a research gap in tackling classification challenges, which remains the core focus of the present study. To evaluate the performance of the proposed reliable jumping-based intelligent classifier in environmental decision-making, we considered ten benchmark datasets spanning various application domains. The numerical results demonstrate that the proposed Reliable Jumping-based intelligent classifier consistently outperforms traditional intelligent classifiers across all studied cases. As a result, the proposed approach proves to be a viable and effective alternative to other intelligent methods in environmental applications.