CONSTRUCTION OF EXPONENTIAL AND LOGISTIC MATHEMATICAL MODELS FOR ESTIMATING THE NUMBER OF POOR POPULATION IN BANGGAI ISLANDS REGENCY

Source abstract

Poverty remains a significant development challenge requiring effective intervention strategies supported by accurate predictive models. This study constructs exponential and logistic mathematical models to estimate the poor population in Banggai Kepulauan Regency, using historical data from 2013 to 2024. The models were constructed through linearization techniques, and parameters were estimated using the least squares method. Both models demonstrated excellent accuracy levels with Mean Absolute Percentage Error (MAPE) values below 10%, where the exponential model achieved MAPE of 2.17% and the logistic model 4.41%. Sensitivity analysis through growth rate parameter variations confirmed model robustness. Projections for 2025–2030 indicate a declining trend in poverty, with the exponential model N(t) = 19,400 × e-0.022574t predicting approximately 13,217 poor people by 2030. These findings provide scientific evidence that the exponential model can serve as a reliable analytical tool for poverty alleviation policy planning in Banggai Kepulauan Regency, enabling more targeted and evidence-based interventions