Trend Prediction of the Number of Feature Test Requests at the Indonesia Digital Test House in 2023-2025 (April) Using the Prophet Machine Learning Model to Support Decision-Making
Muhammad Habibie Rahman(a*), Riser Fahdiran (a), Yudha Pratama Putra (b)

a)Department of Physics, Faculty of Mathematics and Natural Sciences, Universitas Negeri Jakarta
*muhabibie.1005[at]gmail.com
b)Indonesia Digital Test House (IDTH)

Abstract

This study presents a time series analysis and forecasting of feature testing requests at the Indonesia Digital Test House (IDTH) covering the period from January 2023 to April 2025. The objective is to quantitatively analyze historical trends in testing requests and apply the Prophet machine learning model to produce monthly demand forecasts. The methodological approach includes data preprocessing, trend visualization, and model evaluation using statistical metrics such as MAE, RMSE, and MAPE. The results demonstrate the Prophet model^s ability to capture seasonal patterns and long-term trends in the data. Although the model yields a MAPE of 38.32%, the forecasts provide valuable insights to support strategic decision-making, particularly in operational planning and workload management. Additionally, the study produces informative and predictive visualizations to assist the Quality Unit in optimizing laboratory service strategies. This work highlights the practical application of predictive analytics in the domain of telecommunication device testing and contributes to data-driven planning in laboratory operations.

Keywords: Time Series Forecasting, Prophet Model, Predictive Analytics, MAPE, Decision-Making

Topic: Instrumentation and Computational Physics