microsimulation, big data and predictive analyticsdec 20, 2016  · mark birkin professor of spatial...

24
Mark Birkin Professor of Spatial Analysis and Policy, School of Geography, University of Leeds Director, ESRC Consumer Data Research Centre Director, Leeds Institute for Data Analytics Microsimulation, Big Data and Predictive Analytics

Upload: others

Post on 02-Oct-2020

2 views

Category:

Documents


1 download

TRANSCRIPT

Page 1: Microsimulation, Big Data and Predictive AnalyticsDec 20, 2016  · Mark Birkin Professor of Spatial Analysis and Policy, School of Geography, University of Leeds Director, ESRC Consumer

Mark Birkin Professor of Spatial Analysis and Policy, School of Geography, University of Leeds Director, ESRC Consumer Data Research Centre Director, Leeds Institute for Data Analytics

Microsimulation, Big Data and Predictive Analytics

Page 2: Microsimulation, Big Data and Predictive AnalyticsDec 20, 2016  · Mark Birkin Professor of Spatial Analysis and Policy, School of Geography, University of Leeds Director, ESRC Consumer

•  National statistical authorities •  UK Government departments •  Research Institutes •  World-wide data archives

•  Identifying suitable data •  Negotiating access •  Identifying a safe data

access setting

Administrative Data Research Network

Phase1 Phase2 Phase3

Social Media Data & Third Sector Data

Further

announcements soon

ESRCBigDataNetwork

Page 3: Microsimulation, Big Data and Predictive AnalyticsDec 20, 2016  · Mark Birkin Professor of Spatial Analysis and Policy, School of Geography, University of Leeds Director, ESRC Consumer

Multilateral data

sharing

Case studies & academic publications

Metadata & provenance

Business engagement & awareness

raising

Providers

Partners

Prospects

Participants

CDRCDataPartners

Page 4: Microsimulation, Big Data and Predictive AnalyticsDec 20, 2016  · Mark Birkin Professor of Spatial Analysis and Policy, School of Geography, University of Leeds Director, ESRC Consumer

CDRCDataPartners

Page 5: Microsimulation, Big Data and Predictive AnalyticsDec 20, 2016  · Mark Birkin Professor of Spatial Analysis and Policy, School of Geography, University of Leeds Director, ESRC Consumer

HealthSimula=on

Spa$alMicrosimula$on(2011)

GenderxAgexEthnicity

GenderxAgexIllness

GenderxAgexNSSEC

GenderxAgexCarownership

ELSAWave5

HybridMicrosimula$on(2011to2031)

AdjustedProjec$ons

GenderxAgexEthnicity

ETHPOP2011to2031Projec$ons

CHD;stroke;diabetes;cancer;respiratoryillness;

arthri$sanddepression

HazardModel

2011Census

ELSAWaves1to6

Takeaccountoffutureethniccomposi=onofthelocalauthoritypopula=on

ClarkS.,BirkinM.,HeppenstallA.(2014)Subregionales$matesofmorbidi$esintheEnglishelderlypopula$on,Health&Place.

Page 6: Microsimulation, Big Data and Predictive AnalyticsDec 20, 2016  · Mark Birkin Professor of Spatial Analysis and Policy, School of Geography, University of Leeds Director, ESRC Consumer

Health(2)

Page 7: Microsimulation, Big Data and Predictive AnalyticsDec 20, 2016  · Mark Birkin Professor of Spatial Analysis and Policy, School of Geography, University of Leeds Director, ESRC Consumer

Health(3)

Page 8: Microsimulation, Big Data and Predictive AnalyticsDec 20, 2016  · Mark Birkin Professor of Spatial Analysis and Policy, School of Geography, University of Leeds Director, ESRC Consumer

Health (4): App Data

Clockschange Clockschange

Weeklyac$vityreadingsfromtheBountsapp:

Page 9: Microsimulation, Big Data and Predictive AnalyticsDec 20, 2016  · Mark Birkin Professor of Spatial Analysis and Policy, School of Geography, University of Leeds Director, ESRC Consumer

Key:

Transport: Cycling

Blue = commuter cycling potential Green = travel to school cycling potential

Source:hZp://rpubs.com/RobinLovelace/245696

Page 10: Microsimulation, Big Data and Predictive AnalyticsDec 20, 2016  · Mark Birkin Professor of Spatial Analysis and Policy, School of Geography, University of Leeds Director, ESRC Consumer

Transport: Trains

Page 11: Microsimulation, Big Data and Predictive AnalyticsDec 20, 2016  · Mark Birkin Professor of Spatial Analysis and Policy, School of Geography, University of Leeds Director, ESRC Consumer

Transport: Trains

Page 12: Microsimulation, Big Data and Predictive AnalyticsDec 20, 2016  · Mark Birkin Professor of Spatial Analysis and Policy, School of Geography, University of Leeds Director, ESRC Consumer

Journeytowork

Transport: Journey Planning

Page 13: Microsimulation, Big Data and Predictive AnalyticsDec 20, 2016  · Mark Birkin Professor of Spatial Analysis and Policy, School of Geography, University of Leeds Director, ESRC Consumer

Infrastructure

Page 14: Microsimulation, Big Data and Predictive AnalyticsDec 20, 2016  · Mark Birkin Professor of Spatial Analysis and Policy, School of Geography, University of Leeds Director, ESRC Consumer

Infrastructure

Page 15: Microsimulation, Big Data and Predictive AnalyticsDec 20, 2016  · Mark Birkin Professor of Spatial Analysis and Policy, School of Geography, University of Leeds Director, ESRC Consumer

Zuoetal(2014)Geospa$alInforma$onScience,17,3,153-169.

Energy

Page 16: Microsimulation, Big Data and Predictive AnalyticsDec 20, 2016  · Mark Birkin Professor of Spatial Analysis and Policy, School of Geography, University of Leeds Director, ESRC Consumer

Retailing/Consump=on

Page 17: Microsimulation, Big Data and Predictive AnalyticsDec 20, 2016  · Mark Birkin Professor of Spatial Analysis and Policy, School of Geography, University of Leeds Director, ESRC Consumer

Retailing/Consump=on

Page 18: Microsimulation, Big Data and Predictive AnalyticsDec 20, 2016  · Mark Birkin Professor of Spatial Analysis and Policy, School of Geography, University of Leeds Director, ESRC Consumer

RecaponExamples

Page 19: Microsimulation, Big Data and Predictive AnalyticsDec 20, 2016  · Mark Birkin Professor of Spatial Analysis and Policy, School of Geography, University of Leeds Director, ESRC Consumer

“anewmethodofpushingforwardthefron$ersofknowledge,enabledbynewtechnologiesforgathering,manipula$ng,analyzinganddisplayingdata”

TheFourthParadigm…

Thousand years ago, science was empirical, describing natural phenomena

Last few hundred years, theoretical branch, using models, generalizations

Last few decades, a computational branch, simulating complex phenomena Today, data exploration (eScience)

synthesizing theory, experiment and computation with advanced data management and statistics à new algorithms!

(Alex Szalay)

Page 20: Microsimulation, Big Data and Predictive AnalyticsDec 20, 2016  · Mark Birkin Professor of Spatial Analysis and Policy, School of Geography, University of Leeds Director, ESRC Consumer

Predic=veAnaly=cs

•  Dataisgrowing,butwhataboutyourabilitytomakedecisionsbasedonthosehugevolumesofdata?

•  SuccesswilldependonhowquicklyyoucandiscoverinsightsfromallthatdataandusethoseinsightstodrivebeZerac$onsacrossyouren$reorganiza$on

•  That’swherepredic$veanaly$cs,datamining,machinelearninganddecisionmanagementcomeintoplay.

–  Predic$veanaly$cshelpsassesswhatwillhappeninthefuture.–  DatamininglooksforhiddenpaZernsindatathatcanbeusedtopredictfuture

behaviour.Businesses,scien$stsandgovernmentshaveusedthisapproachforyearstotransformdataintoproac$veinsights.

–  Decisionmanagementturnsthoseinsightsintoac$onsthatareusedinyouropera$onalprocesses.

–  (sas.com)

Page 21: Microsimulation, Big Data and Predictive AnalyticsDec 20, 2016  · Mark Birkin Professor of Spatial Analysis and Policy, School of Geography, University of Leeds Director, ESRC Consumer

A Note on Synthesis

•  Twodefini$onsofSynthe.c:– devised,arranged,orfabricatedforspecialsitua$onstoimitateorreplaceusualreali$es

– madebycombiningdifferentsubstances:notnatural

•  Thesecondisrelevantaswellasthefirst!

Page 22: Microsimulation, Big Data and Predictive AnalyticsDec 20, 2016  · Mark Birkin Professor of Spatial Analysis and Policy, School of Geography, University of Leeds Director, ESRC Consumer

Three problems for MSM

•  Calibra$on– andvalida$on?– especiallyinreal-$me

•  Behaviour– Fromdemographicstructuretoac$vi$esandimpacts

•  Predic$veanaly$cs–  robustnessandrelevance

•  Microsimula=onneedsBigData!

Page 23: Microsimulation, Big Data and Predictive AnalyticsDec 20, 2016  · Mark Birkin Professor of Spatial Analysis and Policy, School of Geography, University of Leeds Director, ESRC Consumer

Three problems for Big Data •  Representa$onbias

–  Inferencesforthepopula$on,notthesample

•  Synthesis– Contextisalwayscrucial

•  Privacy,confiden$ality,ownershipandtrust– Conflictbetweengranularityanddisclosure

•  BigDataneedsMicrosimula=on!!

Page 24: Microsimulation, Big Data and Predictive AnalyticsDec 20, 2016  · Mark Birkin Professor of Spatial Analysis and Policy, School of Geography, University of Leeds Director, ESRC Consumer

Conclusions

•  StrongsynergybetweenMSMandBigData•  Dynamicsovermul$plescalesincreasinglyimportant

•  Considersynthe$cpopula$onsassynthesisingaswellassynthesised