Prediction of Birth Rates Using the Naive Bayes Algorithm in the North Sumatra Region

Abstract

The birth rate constitutes a significant metric within the field of demographics. It serves as a crucial element influencing the strategic planning and development of a region, particularly in provinces characterized by notably large populations, such as North Sumatra. The process of forecasting an optimal birth rate necessitates the involvement of multiple agencies and services to effectively devise policies pertaining to health, education, and enhanced infrastructure for the future. This study employs modeling techniques utilizing the Naive Bayes algorithm. This particular algorithm represents a probabilistic classification method within the realm of data mining, aimed at predicting birth rates across all districts and municipalities in North Sumatra, leveraging demographic and socio-economic datasets commencing from the year 2022. The dataset encompasses variables such as population statistics, demographics of women of reproductive age, levels of educational attainment, accessibility to health services, and incidences of poverty, all of which were sourced from the Central Statistics Agency (BPS) over a five-year timeframe. The research methodology is executed through several phases, including data preprocessing, feature selection, partitioning of training and test datasets, and a validation testing process to affirm the reliability of the proposed model. The dataset is partitioned into training and test components utilizing a distribution ratio of 70:30. The outcomes of the proposed model's testing are computed, employing a confusion matrix to derive metrics such as accuracy, precision, recall, and F1 scores. The results yield an accuracy value of 85%, a precision of 87%, and an F1 score of 86%. These findings indicate a favorable outcome in regional mapping and reflect an appropriate birth rate. Subsequently, the results are visualized within a geographic information system (GIS) to elucidate the spatial patterns of the predicted birth rate, thereby facilitating local government interventions in specific areas.