INSIGHT DATA SCIENCE MARTIN GASCON
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
INSIGHT DATA SCIENCE
MARTIN GASCON
HOW DOES IT WORK?
• retrieve info from the GDELT DB • select events based on location• uses a Kernel Density Estimation
The GDELT Project
Events: kill, arrest, kidnap, etc.NumArts: # of articles x eventNumber: # of people affectedGeo location
The tools
The algorithm
The Data (30 GB)
The Database
Global Database of Events, Language and Tone
EXPLORATORY ANALYSIS#
of V
icti
ms
Sundays
Date (Jul-Aug 2015)
It is better to travel on Sundays
EXPLORATORY ANALYSIS
# of Victims
Worldwide victims(06/2015-01/2016)
2 victims
Number of travelers also matters
About me
Martin GasconPhD Nuclear Physics, USC (Spain)Postdoctoral Fellow, Stanford UniversityPhysics Research Fellow, Berkeley Lab
EXTRA SLIDES
Validation - Violent crime score
Information about population for each
Mexican State
Validation - Homicide score
Information about population for each
Mexican State
Hom
icid
e
Validation - Official Data from Mexico Public Safety office
Information about population for each
Mexican State
Validation - Official Data from Mexico Public Safety office
Information about population for each
Mexican State
Validation - Official Data from Mexico Public Safety office
Information about population for each
Mexican State
• advantages and disadvantages of each implementation
• bandwidth
• six kernel forms available in Scikit-learn
EXPLORATORY ANALYSIS
# of Victims
Worldwide victims reflected in at least 4 news
(06/2015-01/2016)
2 victims
• two is the most frequent number of victims
• multiples of ten are frequent too
# of
Vic
tim
s
Sundays
It is better to travel on Sundays
Related Documents
