ARMS: An Attendance-Linked Early Risk Prediction and Academic Record Management Framework for Higher Education

ARMS embeds automated result management alongside an attendance-linked early-warning indicator directly within standard instructor workflows. In synthetic benchmarks, the system identified attendance-driven vulnerabilities overlooked by conventional failure-tally heuristics while achieving ranking performance on par with logistic regression. Our evaluation simultaneously highlighted three operational limitations: a recall-skewed default threshold, an inability to account for historical failure frequency, and diminishing marginal discriminative utility from attendance when performance metrics are already accessible. Future work will curate anonymized institutional datasets encompassing marks, attendance, and formal pass/fail outcomes under institutional ethical clearance to optimize model weights, refine operational cutoffs, and benchmark against tree-based ensembles. Subsequent iterations will introduce a failure-aware composite heuristic alongside planned enhancements: automated study-plan synthesis, parent alert channels, anomaly detection for grading discrepancies, and expanded analytical administrative views.

[1]                    U.S. Department of Education, “Issue brief: Early warning systems.” [Online]. Available: https://www.ed.gov/sites/ed/files/rschstat/eval/highschool/early-warning-systems-brief.pdf [2]                    Y.-H. Chang, F.-C. Chen, and C.-I. Lee, “Developing an early warning system with personalized interventions to enhance academic outcomes for at-risk students in Taiwanese higher education,” Educ. Sci., vol. 15, no. 10, Art. no. 1321, Oct. 2025, doi: 10.3390/educsci15101321. [3]                    C. Romero and S. Ventura, “Educational data mining: A review of the state of the art,” IEEE Trans. Syst., Man, Cybern. C, Appl. Rev., vol. 40, no. 6, pp. 601–618, Nov. 2010. [4]                    E. A. Alyahyan and D. Düştegör, “Predicting academic success in higher education: Literature review and best practices,” Int. J. Educ. Technol. High. Educ., vol. 17, Art. no. 3, 2020, doi: 10.1186/s41239-020-0177-7. [5]                    K. E. Arnold and M. D. Pistilli, “Course Signals at Purdue: Using learning analytics to increase student success,” in Proc. 2nd Int. Conf. Learn.Analytics Knowl. (LAK), Vancouver, BC, Canada, 2012, doi: 10.1145/2330601.2330666. [6]                    D. Bañeres, M. E. Rodríguez, A. E. Guerrero-Roldán, and A. Karadeniz, “An early warning system to detect at-risk students in online higher education,” Appl. Sci., vol. 10, no. 13, Art. no. 4427, Jun. 2020, doi: 10.3390/app10134427. [7]                    L. Cattell and J. Bruch, “Identifying students at risk using prior performance versus a machine learning algorithm,” U.S. Dept. Educ., Inst. Educ. Sci., REL Mid-Atlantic, Rep. REL 2021-126, Sep. 2021. [8]                    Pallets Projects, “Flask documentation.” [Online]. Available: https://flask.palletsprojects.com [9]                    F. Pedregosa et al., “Scikit-learn: Machine learning in Python,” J. Mach. Learn. Res., vol. 12, pp. 2825–2830, 2011. [10]                  L. Breiman, “Random forests,” Mach. Learn., vol. 45, no. 1, pp. 5–32, 2001. [11]                  T. Fawcett, “An introduction to ROC analysis,” Pattern Recognit. Lett., vol. 27, no. 8, pp. 861–874, 2006.