A Hybrid Intelligent Framework for Stress Management: Integrating Machine Learning and Analytical Network Process for Engineering Education Systems

Abstract

Final-year engineering students frequently face severe psychological stress driven by demanding capstone projects and rigorous accreditation standards. This research proposes a novel decision-support framework to objectively evaluate, categorize, and systematically manage academic stress among engineering undergraduates. By integrating automated non-parametric machine learning with a multi-criteria network decision model, the research bridges diagnostic psychological profiling and strategic institutional intervention. Data were gathered from six hundred eleven final-year students across four engineering departments using a validated psychometric questionnaire. The automated classification engine utilizes a nearest-neighbor algorithm based on spatial Euclidean proximity to categorize individual distress severity into depression, anxiety, and stress levels without relying on rigid parametric assumptions. Subsequently, a strategic situational assessment and a multi-perspective personal performance framework were established to translate abstract psychological findings into measurable indicators. Clinical experts and academic domain specialists evaluated these indicators using an analytical network process, incorporating complex interdependencies and feedback loops to compute steady-state priority weights. Empirical results revealed that over ninety percent of respondents experienced at least moderate psychological distress. Crucially, the mathematical prioritization demonstrated that the financial perspective holds the paramount priority weight, followed by internal cognitive resilience, continuous knowledge acquisition, and external social relationships. This proves that financial stability serves as a foundational prerequisite for academic recovery and intellectual execution. The primary contribution of this research lies in converting static psychometric evaluations into a dynamic, clinically validated decision-support tool. Higher education institutions and academic counselors can implement this prioritized framework to allocate institutional resources effectively, design targeted financial relief initiatives, and optimize student well-being alongside academic excellence.