Fuzzy Inference System for Analysing Physical Fitness Metrics of Male-Female Trainee Athletes: De-Fuzzification via Hull and Sigma Scale Analysis

Abstract

A simulation-based and deterministic approach was employed to assess the health-related fitness of upper primary school students through a Fuzzy Inference System (FIS) implemented in MATLAB. Standardized physical assessments were used to gather fitness data, which were then systematically categorized by gender and grade. While statistical metrics such as mean and standard deviation were extracted, inconsistencies and data ambiguities reduced the effectiveness of a strictly deterministic analysis. To overcome these limitations, fuzzy logic was introduced to better manage uncertainty and overlapping patterns in the data. Linguistic variables derived from the Hull and Sigma Scales were incorporated as signal descriptors within the fuzzy framework, improving system interpretability. A triangular membership function was chosen for its balance of computational simplicity and accuracy in classifying fitness levels. Simulation outcomes revealed that the Hull Scale achieved 18% higher consistency in classification compared to the Sigma Scale, highlighting its superior diagnostic performance and potential for identifying health-related fitness trends across diverse student populations. Additionally, optimal input parameters were identified, further enhancing the functionality of decision support systems in school health monitoring. Results confirmed that integrating fuzzy logic with deterministic models leads to a more adaptable and reliable method for assessing student fitness across genders. This hybrid approach can support educators, health professionals, and policymakers in developing more effective, targeted physical wellness interventions. Thus, upper primary students' health-related fitness can be accurately evaluated using a FIS-based system in MATLAB, enhancing performance in youth-focused decision support applications.