Pengembangan Sistem Monitoring Tumbuh Kembang Anak Balita Berbasis Mobile Dengan Explainable Risk-based Prediksi Stunting Dan Umpan Balik Gamifikasi Adaptif

Nutritional deficiencies during early childhood, particularly within the first 1,000 days of life, trigger physical and cognitive growth impairments known as stunting. This study aims to develop "Tumbuh Cerdas," an interactive mobile-based monitoring system designed to facilitate early detection and prevention of stunting. The proposed approach integrates a Random Forest classification model, decision transparency via Explainable AI (XAI) using the SHAP method, and an adaptive rule-based gamification feedback mechanism. Evaluation results demonstrate that the Random Forest model achieves an accuracy rate of 99.18% in classifying child nutritional status based on anthropometric parameters. Functional evaluation through Black Box Testing confirms that the entire system operates according to the designated scenarios, with data synchronized in real-time. The impact of this innovation lies in the creation of a digital health decision support system that is transparent, secure, and educational for both parents and community health workers (kader posyandu) to sustainably reduce stunting prevalence.