Machine learning and game theory is helping fight elephant poaching

Thisarticle is published in collaboration withQuartz.

图1

Between 2009 and 2015,TanzaniaandMozambiquelost more than half of their elephants.

Image:REUTERS/Thomas Mukoya

Writtenby

Ananya

BhattacharyaContributor, Quartz

Published

Friday7 October 2016

Africa’s wildlife is in a constantstate of danger.

Between 2009 and 2015,TanzaniaandMozambiquelost more than halfof their elephants, many of them topoaching forivory smuggling. The decline has propelledAfrican vulture populations, who feed onelephant carcasses, toward extinction too. And attempts atcurtailing poachingand ivory smuggling haven’t helped thedwindling elephant population. InSouth Africa, rhinos are aprized poaching targettoo, for their horns. The attempts tokeep poachers at bay having failed, some conservationists have proposed theexpensive alternative ofairlifting rhinosaway from poaching sites.

Uganda, which remains“heavily implicated”in the illegal ivory trade accordingto the monitoring body CITES, is now testing a more direct way to crack down onthe illegal hunters before they even get to the animals. Using ProtectionAssistant for Wildlife Security (PAWS), a technology combining machine learningand game theory, researchers can predict where poachers may attack and tellrangers where to patrol.

图2

Image: Guardian

“The basic idea is that youhave limited resources, you can’t be everywhere all the time,”UniversityofSouthern Californiaprofessor MilindTambe, who’s leading the initiative, told Quartz. “Where and when should you dopatrol?”

To make their predictions,researchers studied 12 years worth of data collected by rangers, from 2003 to2015, provided by the Wildlife Conservation Society. These included reports ofpast attacks, snare placements, and other illegal activities. The data aren’t perfect,says Tambe: Rangers don’t patrol the entire park, so it’s hard to get acomplete picture. But it’s enough to let a machine learning algorithm makeintelligent guesses about where poachers will strike in future.

When creating patrol routesfor rangers, “we want to randomize our patrols because we ourselves don’t wantto become predictable to the poachers,” Tambe said. That’s where game theorycomes in. It uses mathematical models to evaluate how rational human beingswould act, to then suggest routes that won’t be easily predictable.

The US Coastguard,Transportation Security Administration (TSA), the Federal Air Marshals Service,LA Sheriff’sDepartment, and other organizations have been using Tambe’s AI-game theorycombination technology to randomize their patrols since the early 2000s, hesays. The concept was tailored for wildlife preservation in 2014 and deployedfor testing inMalaysiain mid-2015. The current large-scale Ugandan tests inQueenElizabethNational Parkare backedby US organizations like the National Science Foundation and the Army ResearchOffice.

Rangers using PAWS inUgandahavefound 10 antelope traps and elephant snares in the past month, “a far betterscore card than they could usually expect,”Reuters reported. As robust as the technologymight be in theory, factors like poor mobile internet connections can get inthe way of communicating the results from PAWS that are used to direct rangers’routes. And there’s another threat: Armed poachers are quick topoint their gunsat the rangers.

.