General Population Projection Models with Census Data

This paper applies machine learning methods to generate effective population projections using the American Community Survey (ACS) 5-Year Estimate data. To quantify the error margin in the population projection for each city, conformalized quantile regression, a form of conformal prediction, is used. The modeling process encompasses both statistical machine learning models and deep learning models. The results are presented as visualizations, featuring combined plots of selected statistics for Riverside County in California, which serve as the test dataset.