<![CDATA[Stunting is a chronic nutritional condition that requires rapid and accurate early detection. Conventional assessment methods that rely on the Indonesian Maternal and Child Health (MCH) handbook; commonly referred to as the MCH handbook or manual anthropometric tables often face challenges related to efficiency and susceptibility to human error. The digitization of anthropometric assessment through an Android‑based application offers a promising solution for automating the evaluation of toddlers’ nutritional status. This study aimed to examine the perceived benefits and user acceptance of a machine learning–based Android application designed for stunting detection and management. A descriptive quantitative design was employed. Consecutive sampling was used to recruit individuals who met the inclusion criteria, including the ability to operate the designated Android system, resulting in a sample of 37 respondents. Data were collected using a structured questionnaire with a 1–4 Likert scale. Descriptive statistics were applied to generate frequency distributions, percentages, and mean scores for the indicators of accuracy, usability, utility, and reliability. Findings indicate a high level of user acceptance. Usability received the highest mean score (3.28), followed by utility (3.25) and reliability (3.16). A total of 91.7% of respondents rated the prediction system as accurate, with an average accuracy score of 3.05. Respondents also reported that the application was substantially more efficient and faster than manual assessment methods. In conclusion, the machine learning–based stunting detection application effectively supports users in identifying stunting risk quickly and accurately. Its ease of use and strong perceived validity make it a practical tool for both healthcare workers and the general public, contributing to efforts to accelerate stunting reduction.Keywords: anthropometry; Android-based application; machine learning; stunting]]>
