Fuzzy Logic Based Model for Predicting Neurasthenia in Nigerian University Students

This study developed a fuzzy logic–based predictive model for estimating the likelihood of neurasthenia among students, focusing on interpretable, and non-invasive factors. Risk factors, including academic stress, sleep quality, socioeconomic status, study habits, psychostimulant use, and emotional distress, were identified from literature and expert consultation. These were mapped to linguistic terms and triangular membership functions within a Mamdani fuzzy inference system designed in MATLAB Fuzzy Logic Toolbox (R2024b). A rule base was formulated from expert knowledge, and the system was simulated to evaluate the likelihood of Neurasthenia. The model incorporated six input variables and a three-level output classification: low likelihood, moderate likelihood, and high likelihood of having neurasthenia. Simulation results based on the 729 rules, indicated that 25% of cases were classified as low likelihood, 40% as moderate likelihood and 35% as high likelihood of Neurasthenia. These percentages reflect the distribution across a range of input combinations where unfavorable conditions, such as high academic stress, poor sleep, frequent stimulant use, and severe emotional distress, consistently produced a high-likelihood output, while positive conditions, such as effective study habits, good sleep, and low academic stress, resulted in low likelihood of Neurasthenia. The MATLAB desktop application successfully implemented the model, providing an intuitive interface for prediction. In conclusion. this study demonstrated the effectiveness of fuzzy logic in predicting neurasthenia risk by modeling uncertainty and enhancing interpretability. The tool provides a foundation for early detection and awareness in academic institutions, with potential applications in student health support systems.