美赛数学建模2010____B优秀论文

美赛数学建模优秀论文

Centroids,Clusters,andCrime129

Centroids,Clusters,andCrime:

AnchoringtheGeographicPro lesofSerialCriminals

AnilS.Damle

ColinG.West

EricJ.Benzel

UniversityofColorado–Boulder

Boulder,CO

Advisor:AnneDougherty

Abstract

Aparticularlychallengingproblemincrimepredictionismodelingthebehaviorofaserialkiller.Since ndingassociationsbetweenthevictimsisdif cult,wepredictwherethecriminalwillstrikenext,insteadofwhom.Suchpredictingofacriminal’sspatialpatternsiscalledgeographicpro ling.Researchshowsthatmostviolentserialcriminalstendtocommitcrimesinaradialbandaroundacentralpoint:home,workplace,orotherareaofsigni cancetothecriminal’sactivities(forexample,apartoftownwhereprostitutesabound).These“anchorpoints”providethebasisforourmodel.Weassumethattheentiredomainofanalysisisapotentialcrimespot,movementofthecriminalisuninhibited,andtheareainquestionislargeenoughtocontainallpossiblestrikepoints.Weconsiderthedomainametricspaceonwhichpredictivealgorithmscreatespatiallikelihoods.Addition-ally,weassumethattheoffenderisa“violent”serialcriminal,sinceresearchsuggeststhatserialburglarsandarsonistsarelesslikelytofollowspatialpatterns.

Therearesubstantialdifferencesbetweenoneanchorpointandseveral.Wetreatthesingle-anchor-pointcase rst,takingthespatialcoordinatesofthecriminal’slaststrikesandthesequenceofthecrimesasinputs.Estimatingthepointtobethecentroidofthepreviouscrimes,wegeneratea“likelihoodcrater,”whereheightcorrespondstothelikelihoodofafuturecrimeatthatlocation.Forthemultiple-anchor-pointcase,weuseacluster- ndingandsortingmethod:Weidentifygroupingsinthedataandbuildalikelihoodcrateraroundthecentroidofeach.Eachclusterisgivenweightaccordingtorecencyandnumberofpoints.Wetestsinglepointvs.multiplepointsby

cCopyright2010byCOMAP,Inc.Allrightsreserved.TheUMAPJournal31(2)(2010)129–148.

Permissiontomakedigitalorhardcopiesofpartorallofthisworkforpersonalorclassroomuseisgrantedwithoutfeeprovidedthatcopiesarenotmadeordistributedforpro torcommercialadvantageandthatcopiesbearthisnotice.Abstractingwithcreditispermitted,butcopyrightsforcomponentsofthisworkownedbyothersthanCOMAPmustbehonored.Tocopyotherwise,torepublish,topostonservers,ortoredistributetolistsrequirespriorpermissionfromCOMAP.

你可能喜欢

  • 数学建模美赛优秀论文
  • 模拟退火算法
  • 粒子群优化算法
  • 人工神经网络模型
  • BP神经网络原理
  • 综合评价法

美赛数学建模2010____B优秀论文相关文档

最新文档

返回顶部