DWP Sparkle Talk

Simulating Society with Agent-Based Modelling and Microsimulation


Nick Malleson

University of Leeds, UK

n.s.malleson@leeds.ac.uk


Slides available at:
https://urban-analytics.github.io/dust/presentations.html

Overview

Introduction to ABM

ABM Example: Simulating daily mobility

Introduction to Microsimulation

Microsimulation examples

Simulating implications for tax policies

Future ageing

Discussion

Diagram of regression

Introduction to ABM

Aggregate v.s. Individual

'Traditional' modelling methods work at an aggregate level, from the top-down

E.g. Regression, spatial interaction modelling, location-allocation, etc.

Aggregate models work very well in some situations

Homogeneous individuals

Interactions not important

Very large systems (e.g. pressure-volume gas relationship)

Diagram of regression

Introduction to ABM

Aggregate v.s. Individual

But they miss some important things:

Low-level dynamics, i.e. “smoothing out” (Batty, 2005)

Interactions and emergence

Individual heterogeneity

Unsuitable for modelling complex systems

Diagram of the sims

Introduction to ABM

Systems are driven by individuals

(cars, people, ants, trees, whatever)

Bottom-up modelling

An alternative approach to modelling

Rather than controlling from the top, try to represent the individuals

Account for system behaviour directly

Autonomous, interacting agents

Represent individuals or groups

Situated in a virtual environment

A termite mound.
Attribution: JBrew (CC BY-SA 2.0).

Why ABM?

Emergence

"The whole is greater than the sum of its parts." (Aristotle?)

Simple rules → complex outcomes

E.g. who plans the air-conditioning in termite mounds?

Hard to anticipate, and cannot be deduced from analysis of an individual

ABM uses simulation to (try to) understand how macro-level patterns emerge from micro-level behaviours

Why ABM?

Better Representations of Theory

Example: Crime theories emphasise importance of ...

Individual behaviour (offenders, victims, guardians)

Individual geographical awareness

Environmental backcloth

Why ABM?

Better Representations of Space

Example of GIS data

Micro-level environment is very important

Can richly define the space that agents inhabit

More Natural Description of a System

Describe the entities directly, rather than using aggregate equations

Why ABM?

History of the Model Evolution

Rather than returning a single result, the model evolves

The evolution itself can be interesting

Analyse why certain events occurred

Diagram illustrating different ABM applications by behavioural and system/environment complexity

Modelling agent behaviours

Many behaviours are hard / impossible to model

Choose those that are the most important. Cannot include everything!

Some can be very simple - e.g. threshold-based rules (Kennedy, 2012)

IF hunger IS ABOVE hunger_threshold THEN search_for_food
OTHERWISE do_something_else

These are the most common (Birks et al. 2012, 2013; Dray et al. 2008; Groff 2007a,b; Hayslett-McCall, 2008)

More advanced cognitive frameworks exist

Beliefs, Desires, Intentions (Bratman et al., 1988)

PECS (Schmidt, 2000).

ABM Predictive Example

Awareness space test

Agent-Based Modelling - Difficulties

Will I play with the truck, or the duck?

(actually he played with his trains...)

Tendency towards minimal behavioural complexity

Stochasticity

Computationally expensive (not amenable to optimisation)

Complicated agent decisions, lots of decisions, multiple model runs

Modelling "soft" human factors

Need detailed, high-resolution, individual-level data

Individual-level data

ABM Example:
Simulating Urban Mobility

Motivation: better models of daily urban dynamics by combining diverse data and simulation

Simulating Urban Flows (surf) and Data Assimilation for Agent-Based Modelling (dust) projects

Crols, T., and N. Malleson (2019) Quantifying the Ambient Population Using Hourly Population Footfall Data and an Agent-Based Model of Daily Mobility. GeoInformatica DOI: 10.1007/s10707-019-00346-1.

Malleson, N., K. Minors, Le-Minh Kieu , J. A. Ward , A. West and A. Heppenstall (2020) Simulating Crowds in Real Time with Agent-Based Modelling and a Particle Filter. Journal of Artificial Societies and Social Simulation (JASSS) 23 (3). http://jasss.soc.surrey.ac.uk/23/3/3.html DOI: 10.18564/jasss.4266

Simulating Urban Mobility

Wi-Fi footfall counters. Case study: Otley, West Yorkshire

The surf model environment in Otley The surf model environment in Otley (zoom in)

Simulating Urban Mobility

Diagram illustrating SURF method

Simulating Urban Mobility: Results 1

Initial results from the surf model

Simulating Urban Mobility: Results 2

Results after including commuter agents

Next steps: real-time calibration

Diagram of dynamic data assimilation and an ABM

Data Assimilation for Agent-Based Modelling (dust.leeds.ac.uk)

Microsimulation v.s. ABM

Confusing terminology!

(Spatial) Microsimulation (aka 'Synthetic Population Generation')

"merging of census and social survey data to simulate a population of individuals within households" (Ballas et al., 2018).

Dynamic Microsimulation

Microsimulation plus time (O'Donoghue, 2018).

Simulate the behaviour of micro-units over time

Microsimulation v.s. ABM

So what is the difference?

Arguably comes down to interactions

If the individual units can interact, then it's an ABM

If they can't interact, then it's a microsimulation

Also, microsimulations tend to be empirical. ABMs can be abstract / hypothetical / thought experiments.

But, on the whole, not particularly important what you call your model

Microsimulation Example 1:

Future Elderly Model (FEM)

Background: understanding population ageing is key for planning health and welfare systems

English Longitudinal Study of Ageing (ELSA)

Markov microsimulation model

Generates input populations and transition probabilities from ELSA

Projects the population forward in time

Estimates disease prevalence and economic outputs

Scenario investigation through modification of transition probabilities

Luke Archer, Nik Lomax, Bryan Tysinger (in press)

Microsimulation Example 1:

Future Elderly Model (FEM)

Graph showing reduced lung disease in non-smoking scenario
Luke Archer, Nik Lomax, Bryan Tysinger (in review). A Dynamic Microsimulation Model for Ageing and Health in England: The English Future Elderly Model

Prevalence of lung disease in a reduced smoking scenario

Future work:

assess the burden of chronic disease

evaluate the impact of early life experiences on later-life health

investigate the relationship between socio-economic factors and the quality and quantity of life

quantify the potential impact for interventions targeting key risk factors

Microsimulation Example 2

Simulating Tax Policies

Old, Single and Poor: Using Microsimulation and Microdata to Analyse Poverty and the Impact of Policy Change among Older Australians (Tanton et al., 2009)

Creates a synthetic population using national census and smaller surveys

Examines national, spatial impacts of pension age change on older, single individuals

Microsimulation Example 2

Simulating Tax Policies

Map of distribution of change in poverty rates
Tanton, Robert, Vidyattama, Yogi, McNamara, Justine, Vu, Quoc Ngu and Harding, Ann, (2009), Old, Single and Poor: Using Microsimulation and Microdata to Analyse Poverty and the Impact of Policy Change among Older Australians*, Economic Papers, 28, issue 2, p. 102-120.

Summary

Introduction to ABM

ABM Example: Simulating daily mobility

Introduction to Microsimulation

Microsimulation examples

Simulating implications for tax policies

Future ageing

Now: discussion & questions

DWP Sparkle Talk

Simulating Society with Agent-Based Modelling and Microsimulation


Nick Malleson

University of Leeds, UK

n.s.malleson@leeds.ac.uk


Slides available at:
https://urban-analytics.github.io/dust/presentations.html