Upcoming SlideShare
×

# Система предсказания преступлений

655 views

Published on

Система предсказания преступлений (Santa Cruz Police Department, Калифорния)

Published in: Education, Travel
0 Likes
Statistics
Notes
• Full Name
Comment goes here.

Are you sure you want to Yes No
• Be the first to comment

• Be the first to like this

Views
Total views
655
On SlideShare
0
From Embeds
0
Number of Embeds
4
Actions
Shares
0
2
0
Likes
0
Embeds 0
No embeds

No notes for slide

### Система предсказания преступлений

1. 1. George Mohler                George Tita     SCU Mathematics                UCI Criminology Jeff Brantingham Santa Cruz Police DeptUCLA Anthropology Los Angeles Police Dept  UC MaSC Mathematical And Simulation Modeling of Crime
2. 2.   Certain crime types show an elevated risk of crime  following an event
3. 3.   Retaliation is a cause of inter‐gang violence   Burglars return to the same, or neighboring,  house in the days/weeks after a crime  “I always go back [to the same places] because, once you been there, you know just about when you been there before and when you can go back. An every time I hit a house, it’s always on the same day [of the week] I done been before cause I know there ain’t nobody there. “ (Subject No. 51) Wright and Decker Burglars on the Job
4. 4. Hawkes Processmodel adaptedfrom seismology
5. 5.   Each day ﬁt the Hawkes Process to past data  to estimate risk level across the city   Direct patrols to highest risk areas and times  of day   Next day recalibrate model and repeat    As highest risk areas change, the model and  patrols adapt
6. 6. Prototype Software wI GoogleMaps Interface ~ University of CabnUo Hwy G) california Santa Cruz • Harvey We1>t Park ~ ~ Soquel Ave TWin Lakes e ry co .9. Lagoon Park ~ "",....f ",Y ~ • ~ Q <s- , If) -< ~ • atural BrIdges Slele Beach en ~ u" ~Or L i~ouse • Field Slate Beach
7. 7. P= 0.0207 P=0.0199 Pveh = 0.702, Pres = 0 .297 Pveh = 1, Pres = O TW : 13 : 00, 15 : 00 TW : 12 : 00, 13 : 00 200 Laurel 51P is probability of one or more crimes in 24 hour time periodPveh, Pres are conditional probabilities of vehicular and residential crimeTW gives start times of 2 highest risk 1-hour time windows
8. 8.   Patrol eﬃciency:  quantitative approach to  placing patrols where crime is located   Reduction in crime rate in ﬂagged areas   Reduction in overall crime rate
9. 9. UC MaSC Mathematical and Simulation Modeling of Crime  NSF Human Social Dynamics Program   NSF Division of Mathematical Sciences   US DoD,  ARO, AFOSR, ONR   Santa Clara University, UCLA, UC Irvine   LAPD, Long Beach PD, Santa Cruz PD publications: paleo.sscnet.ucla.edu