Amir Ali, S.Kom, M.Kom, a lecturer at the Dr. Soetomo Hospital Foundation Health College (STIKES) successfully earned his doctorate after defending his dissertation at the doctoral defense of the Information Systems Doctoral Study Program (DSI) at Diponegoro University (Undip) on Wednesday, May 13, 2026.
Dr. Amir Ali, S.Kom, M.Kom conducted his dissertation research entitled "Stunting Prediction Modeling Based on Maternal Pregnancy Examination History and Toddler Anthropometric Data Using Random Forest Integrated with a Geographic Information System."
The examining team included Prof. Ir. Mochamad Agung Wibowo, M.M., M.Sc., Ph.D. (Chairman/Dean of the Graduate School), Prof. Dr. Ir. R. Rizal Isnanto, S.T., M.M., M.T., IPU., ASEAN Eng (Secretary of the Defense/Head of the Information Systems Doctoral Study Program), Prof. Dr. Ir. Qomariyatus Sholihah, Amd.Hyp, S.T., M.Kes., IPU., ASEAN M.Si (External Examiner/Brawijaya University), Prof. Dr. Rahmat Gernowo, M.Si, Ir. Mochammad Facta, S.T., M.T., Ph.D, Prof. Dr. Mundakir, S.Kep., Ns., M.Kep. FISQua, (Co-Promoter/Faculty of Health Sciences, Umsura), and Prof. Dr. Ir. Purwanto, DEA (Promoter).
According to Dr. Amir Ali, stunting in toddlers is a persistent nutritional problem in Indonesia. The 2022 Indonesian Nutritional Status Survey (SSGI) reported that the stunting rate in Sidoarjo Regency increased from 14.8% to 16.1%. According to the 2022 Sidoarjo Regency Profile Book, stunting still accounts for 5.8% of all toddlers measured for height.
"Furthermore, the lack of system and data integration between the e-PPGBM application and the SI-Cantik application owned by the Sidoarjo Regency Health Office has limited information regarding stunting incidence. Therefore, to achieve the government's target of 14% stunting prevalence by 2024, action is needed to reduce stunting prevalence in toddlers," said Dr. Amir Ali.
"The purpose of this study is to create a stunting prediction model based on toddler anthropometric data and maternal pregnancy check-up history using a classification algorithm approach. This model uses the Random Forest algorithm by preprocessing the dataset using the SMOTE method where the evaluation of model performance uses an evaluation matrix. The random forest algorithm was chosen because of its ability to improve accuracy on missing data, prevent errors, and store data efficiently and improve performance in classification models. The SMOTE method was chosen because this method is useful for balancing the distribution of the number of data samples in the minority class by selecting these data samples so that the number is equal to the number of data samples in the majority class.
