Introduction to Spatial Microsimulation - PowerPoint Presentation

Slide1

Introduction to Spatial Microsimulation

Dr Kirk HarlandSlide2

What is a Spatial

Microsimulation

?Static Spatial MicrosimulationDeterministic ReweightingConditional ProbabilitiesSimulated AnnealingDynamic Microsimulation

This

SessionSlide3

What is Spatial Microsimulation

There are two types of Spatial

MicrosimulationStatic spatial microsimulation - creates a micro-level population from aggregate dataDynamic spatial microsimulation – moves a population through space and timeSlide4

Static Spatial Microsimulation

Static spatial microsimulation synthesises individual level populations from aggregate information

Does not move the population through space or timeAlternative approach to joining two datasets spatially where no join is apparent, many health examples includingobesity (Smith et al., 2009)diabetes (Smith et al., 2005)smoking prevalence (Tomintz and Clarke, 2008)Slide5

Static Spatial Microsimulation

Several different static microsimulation methods

Deterministic reweighting – large iterative proportional fitting algorithmConditional probabilities – calculates the probability of a person appearing in a zone give there characteristicsSimulated annealing – combinatorial optimisation algorithm originally designed to simulate the cooling properties of metalsSlide6

Static Spatial Microsimulation

But they all attempt to do the same thing

Turn a selection of aggregate constraint tables Into an individual level population allocated to spatial areasSlide7

Static Spatial Microsimulation

While minimising the difference between the distribution of the constraint table attributes for each zone and the distribution of the attributes aggregated from the synthesised population…

Zones

Zones

Male – gender constraint counts

Male – gender

synthesised

population countsSlide8

Static Spatial Microsimulation

Fit statistic used is normally Total Absolute Error (TAE)

TAE = ∑i∑j|Tij – Eij|WhereTij is the sum of the observed counts for the cell ijEij is the sum of the expected counts for the cell

ij

Williamson et al 1998Slide9

Static Spatial

Microsimulation – Deterministic Reweighting

A very big iterative proportional fitting algorithmStage 1 – calculate weights for each individualSmith et al 2009Slide10

Static Spatial

Microsimulation – Deterministic Reweighting

Stage 2 - proportionally fit each weight to the population

Smith et al 2009Slide11

Static Spatial

Microsimulation – Deterministic Reweighting

Iterate over the reweighting process until: the fit statistic does not improve any furtherA threshold set on the fit statistic to indicate convergence is reachedMove to next zoneThis algorithm has been widely used in health studies.Slide12

Static Spatial

Microsimulation – Conditional Probabilities

Birking and Clarke 1988Stage 1 – calculate conditional probabilities for all possible combinations of individualsSlide13

Static Spatial

Microsimulation – Conditional Probabilities

Birking and Clarke 1988Stage 2 – Assign synthetic characteristics applying conditional probabilitiesSlide14

Static Spatial

Microsimulation

– Conditional ProbabilitiesBirking and Clarke 1988Stage 3 – Constrain weights to constraint table distributionsSlide15

Static Spatial

Microsimulation

– Conditional ProbabilitiesBirking and Clarke 1988Stage 4 – Calculate TAEStage 5 – Iterate over previous stages until no further reduction in TAEStage 6 – Move to next zoneParticular strength of the algorithm is that it does not require an input populationSlide16

Static Spatial

Microsimulation

– Simulated Annealingsample population

constraint 1

constraint n…

synthetic population

zone x

aggregation 1

aggregation n…

calculate fitness - TAE

Harland et al. 2012Slide17

A combinatorial

optimisation

algorithm well suited to static spatial microsimulation…Accurate, produces good results because it can take backwards stepsComputationally intensive so care needed when implementing code

Static Spatial

Microsimulation

– Simulated Annealing

Harland et al. 2012Slide18

What do we mean by taking backwards steps?

Crossing the valley between say point A to reach point B

Static Spatial

Microsimulation

– Simulated AnnealingSlide19

Comparing the Approaches

Harland et al 2012

Not any more…Slide20

Dynamic Spatial Microsimulation

Takes a population, whether

synthesised or real world data, and moves it through space and timeUses derived probabilities to determine outcomes for individuals at each time-stepIndividuals can typicallyDieBe bornMigrateGet marriedGet divorced… and any number of other actions for which probabilities can be derivedSlide21

Dynamic Spatial Microsimulation

Time step 0

Time step 1

Time step 2

Transition matrices

Transition matricesSlide22

Dynamic Spatial Microsimulation

Seems simple…

Idea is simple but many complicating factorsNumber of transitional probabilities dependent on number of attributesBirth, death, migration, etc… not ubiquitous across zonesDerivation of probabilities become more complex and burdensome than the modelling process.With large populations over longer time periods models can take time to setup and run, causing difficulties with calibration and evaluationSlide23

A Word on

M

odel EvaluationAll too often not dealt with sufficiently in the literature.Williamson and Voas (1998) presented work into model evaluation and assessmentHarland et al. (2012) examined three different model approaches evaluating the algorithm performanceEvaluation of large models is very difficult and time consuming but for reliable results it needs to be doneDifferent levels of statistics provide information about different areas of the modelCell level – fine grained (often not presented)Attribute level – medium detail (often not presented)

Constraint level – high level model assessmentSlide24

Microsimulation

Vs Agent-Based ModellingGreat deal of similarity between the two approachesBoth operate at the individual levelDynamic microsimulation moves individuals through time as does ABMCould argue for simple behaviour in dynamic microsimulation

Both are very data hungry

Also several differences

ABMs are enhanced by interaction of individuals with their environment

Behaviour

in ABM not restricted to simple transitional probabilities

ABMs cannot handle the volumes of data… yet!Slide25

Static spatial

microsimulation

synthesises an individual level population from aggregate dataA variety of approaches have been used for static spatial microsimulation -iterative reweighting

-statistical probabilities

-combinatorial optimisation

All have there benefits and there drawbacks…

SummarySlide26

Dynamic

microsimulation

moves a population through timeHas similarities to ABM but also major differencesStatic spatial microsimulation may have a role to play with both approachesOne major complicating factor for dynamic

microsimulation

is the derivation of transitional probabilities…

SummarySlide27

References

Ballas

, D., Clarke, G., Dorling, D., Eyre, H., Thomas, B., and Rossiter, D.(2005) SimBritain: a spatial microsimulation approach to population dynamics. Population, Space and Place 11, 13–34.Birkin, M. & Clarke, M. (1988). SYNTHESIS - a synthetic spatial information system for urban and regional analysis: methods and examples''

Environment and Planning A, 20,

1645 -1671.

Harland K.,

Heppenstall

A. J., Smith D., and

Birkin

, M. (2012)

Creating Realistic Synthetic Populations at Varying Spatial Scales: A Comparative Critique of Population Synthesis Techniques

Journal of Artificial Societies and Social Simulation

15 (1) 1

Smith M D, Clarke P G, Ransley J, Cade J. (2005). Food Access and Health : A Microsimulation Framework for Analysis. Studies in Regional Science

. 35(4). 909 – 927Smith D M, Clarke G P, Harland K, (2009), Improving the synthetic data generation process in spatial microsimulation models. Environment and Planning A 41(5) 1251 – 1268 Tomintz MN, GP Clarke, (2008) The geography of smoking in Leeds: estimating individual smoking rates and the implications for the location of stop smoking services. Area 40(3): 341-353

Williamson, P., Birkin, M., & Rees, P.H. (1998). The estimation of population microdata by using data from small area statistics and samples of anonymised records. Environment and Planning A, 30, 785-816. Wu, B., Birkin

, M. and Rees P. (2008) A spatial microsimulation model with student agents. Computers, Environment and Urban Systems, 32 (6). pp. 440–453