|
Activity Number:
|
309
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Tuesday, August 4, 2009 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Government Statistics
|
| Abstract - #303966 |
|
Title:
|
Constructing Synthetic Samples Using Simulated Annealing
|
|
Author(s):
|
Glen Meeden and Hua Dong*+
|
|
Companies:
|
The University of Minnesota and Amgen, Inc.
|
|
Address:
|
1418 Bellevue Ave, Burlingame, CA, 94010,
|
|
Keywords:
|
finite sampling ; simulated annealing ; census data
|
|
Abstract:
|
We consider the problem of constructing a synthetic sample for a population which we cannot sample from but for which we know the population values for a set of population quantities. Our approach is to combine subsamples from samples from two similar populations to construct the synthetic sample. We show that on the average such synthetic samples behave very much like actual samples from the population of interest. The synthetic sample is obtained by solving an optimization problem, where the known population parameters will be used as constraints. The optimization is achieved through a simulated annealing algorithm. Simulation studies are presented to demonstrate the effectiveness of our approach. We will discuss applying this approach to constructing a synthetic 5% sample for the 1890 US census for which the complete records where destroyed.
|
