- 2.1. Summary of the notation used for multiway tables and IPF.
- 3.1. Sample sizes of some data sources used for synthesis, at different
levels of geography
- 3.2. An example household containing unusual family structure.
- 3.3. Overview of Person attributes, showing the number of
categories for the attributes in each data source.
- 3.4. Overview of Census Family attributes, showing the number of
categories for the attributes in each data source.
- 3.5. Overview of Household/Dwelling Unit attributes, showing the number of
categories for the attributes in each data source.
- 3.6. The contents of the SC86B01 summary tables: population by sex,
age and highest level of schooling.
- 3.7. Series of log-linear models to test for association between
gender, age and highest level of schooling in SC86B01 table.
- 3.8. Series of log-linear models testing association in SC86B01,
including geography.
- 3.9. Series of log-linear models testing association in SC86B01
relative to PUMS.
- 4.1. Illustration of relationship between sparsity and number of
dimensions/cross-classification attributes.
- 4.2. Comparison of memory requirements for implementations of an
agent synthesis procedure using a complete array or a sparse list.
- 4.3. Format of a sparse list-based data structure for Iterative
Proportional Fitting
- 4.4. Relationship between unknown true count and the randomly rounded
count published by the statistical agency.
- 5.1. Attributes and number of categories used during IPF fitting of
three agent types.
- 5.2. Summary of all attributes that are shared between agents to define
and constrain relationships.
- 5.3. Computation time for the different stages of the synthesis
procedure on a 1.5GHz computer for the Toronto Census Metropolitan Area.
- 6.1. Comparison of
and SRMSE statistics for validation.
- 6.2. Design and results of experiments I1-I10, testing goodness-of-fit
of IPF under varying amounts of input data.
- 6.3. Design and results of experiments R1-R4, testing
goodness-of-fit after using different methods to deal with random rounding.
- 6.4. Design and results of experiments M0-M2, testing
goodness-of-fit after applying different Monte Carlo methods.
- B.1. Validation tables used to evaluate the goodness-of-fit of
synthetic population, with the cell count in parentheses.
- B.2. Detailed results of experiments I1-I10, testing goodness-of-fit
of IPF under varying amounts of input data.
