Synthetic population of Greater Jakarta: An iterative proportional updating approach

Agent based simulation frameworks require synthetic populations to simulate traveller’s activity travel patterns. Various methods exist for population synthesis with different advantages and disadvantages. For the Greater Jakarta area, a synthetic population has previously been generated using Bayesian Network (BN) approach and Generalized Raking (GR) multilevel IPF. This approach used very large, city level marginal to fit the population and only captured individual control totals. Here we generate a synthetic population using Iterative Proportional Updating approach to match both individual and household level controls of the population of smallest available census zones, called subdistricts, in Greater Jakarta region, using census-based marginal distributions and existing sample surveys