This research program leverages large-scale survey data from the Young Adult Fertility and Sexuality Study 5 (YAFS5) of the University of the Philippines Population Institute to understand and address urgent mental health challenges in youth and young adults across ASEAN. It combines statistical modeling, machine learning, and AI-driven tools to uncover factors linked to suicidal attempts, develop actionable recommendations on depression, and design a GenAI chatbot to support student mental well-being.
Visualizing trends in mental well-being and predictive modeling performance using anonymized data patterns.
Dataset: YAFS5 youth data (Philippines, regional comparisons).
Methods: supervised learning (logistic regression, gradient boosting, neural networks), fairness/bias evaluation, interpretable models (SHAP/LIME).
Outcomes: identification of high-risk subgroups, risk prediction models.
Comparative analysis of prevalence, severity, and correlates of depression.
Use of regression modeling, clustering, and cross-country comparisons.
Emphasis on cultural and socioeconomic determinants.
Generative AI chatbot trained on student wellbeing dialogues, mental health resources, and local language corpora.
Safety layers: crisis detection, escalation to human support, bias and toxicity filtering.
Evaluation: pilot testing, user satisfaction surveys, clinical safety assessments.