Predicting poaching for wildlife Protection

Wildlife species such as tigers and elephants are under the threat of poaching. To combat poaching, conservation agencies (“defenders”) need to 1) anticipate where the poachers are likely to poach and 2) plan effective patrols. We propose an anti-poaching tool CAPTURE (Comprehensive Anti-Poaching tool with Temporal and observation Uncertainty REasoning), which helps the defenders achieve both goals. CAPTURE builds a novel hierarchical model for poacher-patroller interaction. It considers the patroller's imperfect detection of signs of poaching, the complex temporal dependencies in the poacher's behaviors, and the defender's lack of knowledge of the number of poachers. Further, CAPTURE uses a new game-theoretic algorithm to compute the optimal patrolling strategies and plan effective patrols. This paper investigates the computational challenges that CAPTURE faces. First, we present a detailed analysis of parameter separation and cell abstraction, two novel approaches used by CAPTURE to efficiently learn the parameters in the hierarchical model. Second, we propose two heuristics—piecewise linear approximation and greedy planning—to speed up the computation of the optimal patrolling strategies. In this paper, we discuss the lessons learned from using CAPTURE to analyze real-world poaching data collected over 12 years in Queen Elizabeth National Park in Uganda.

[1]  Tuomas Sandholm,et al.  Game theory-based opponent modeling in large imperfect-information games , 2011, AAMAS.

[2]  Bo An,et al.  Deploying PAWS: Field Optimization of the Protection Assistant for Wildlife Security , 2016, AAAI.

[3]  Shyam Varan Nath Crime Pattern Detection Using Data Mining , 2006, 2006 IEEE/WIC/ACM International Conference on Web Intelligence and Intelligent Agent Technology Workshops.

[4]  Hugo Jachmann Illegal wildlife use and protected area management in Ghana , 2008 .

[5]  J. Andrew Royle,et al.  ESTIMATING SITE OCCUPANCY RATES WHEN DETECTION PROBABILITIES ARE LESS THAN ONE , 2002 .

[6]  David L. Roberts,et al.  Devaluing rhino horns as a theoretical game , 2016 .

[7]  V. S. Subrahmanian,et al.  APE: A Data-Driven, Behavioral Model-Based Anti-Poaching Engine , 2015, IEEE Transactions on Computational Social Systems.

[8]  Milind Tambe,et al.  When Security Games Go Green: Designing Defender Strategies to Prevent Poaching and Illegal Fishing , 2015, IJCAI.

[9]  Helen M. Regan,et al.  The conservation game , 2011 .

[10]  Michael H. Bowling,et al.  Bayes? Bluff: Opponent Modelling in Poker , 2005, UAI.

[11]  C. Forsyth Chasing and catching bad guys': the game warden's prey , 1993 .

[12]  Walter A. Kosters,et al.  Data Mining Approaches to Criminal Career Analysis , 2006, Sixth International Conference on Data Mining (ICDM'06).

[13]  Heribert Hofer,et al.  MODELING THE SPATIAL DISTRIBUTION OF THE ECONOMIC COSTS AND BENEFITS OF ILLEGAL GAME MEAT HUNTING IN THE SERENGETI , 2008 .

[14]  Milind Tambe,et al.  "A Game of Thrones": When Human Behavior Models Compete in Repeated Stackelberg Security Games , 2015, AAMAS.

[15]  A. Lemieux,et al.  Poaching in Uganda: Perspectives of Law Enforcement Rangers , 2015 .

[16]  S. Eliason Wildlife Crime: Conservation Officers' Perceptions of Elusive Poachers , 2008 .

[17]  Rong Yang,et al.  Adaptive resource allocation for wildlife protection against illegal poachers , 2014, AAMAS.

[18]  Xiao-Li Meng,et al.  Maximum likelihood estimation via the ECM algorithm: A general framework , 1993 .

[19]  Radford M. Neal Pattern Recognition and Machine Learning , 2007, Technometrics.

[20]  D. McFadden Conditional logit analysis of qualitative choice behavior , 1972 .

[21]  Milind Tambe Security and Game Theory - Algorithms, Deployed Systems, Lessons Learned , 2011 .

[22]  Vincent Conitzer,et al.  Complexity of Computing Optimal Stackelberg Strategies in Security Resource Allocation Games , 2010, AAAI.

[23]  Nicola Basilico,et al.  Automated Abstractions for Patrolling Security Games , 2011, AAAI.

[24]  Yevgeniy Vorobeychik,et al.  Computing Randomized Security Strategies in Networked Domains , 2011, Applied Adversarial Reasoning and Risk Modeling.

[25]  Amos Azaria,et al.  Analyzing the Effectiveness of Adversary Modeling in Security Games , 2013, AAAI.

[26]  Rong Yang,et al.  Computing optimal strategy against quantal response in security games , 2012, AAMAS.

[27]  R. Hilborn,et al.  Effective Enforcement in a Conservation Area , 2006, Science.

[28]  Alta de Waal,et al.  An expert-driven causal model of the rhino poaching problem , 2017 .

[29]  Nicola Basilico,et al.  Leader-follower strategies for robotic patrolling in environments with arbitrary topologies , 2009, AAMAS.

[30]  Matthew Linkie,et al.  EDITOR'S CHOICE: Safeguarding Sumatran tigers: evaluating effectiveness of law enforcement patrols and local informant networks , 2015 .

[31]  Milind Tambe,et al.  CAPTURE: A New Predictive Anti-Poaching Tool for Wildlife Protection , 2016, AAMAS.

[32]  Greg L. Warchol,et al.  Policing the wilderness: A descriptive study of wildlife conservation officers in South Africa , 2012 .

[33]  H. Jachmann,et al.  Elephant poaching and law enforcement in the central Luangwa Valley, Zambia , 1997 .

[34]  R Critchlow,et al.  Spatiotemporal trends of illegal activities from ranger-collected data in a Ugandan national park. , 2015, Conservation biology : the journal of the Society for Conservation Biology.

[35]  Milind Tambe,et al.  Keeping Pace with Criminals: Designing Patrol Allocation Against Adaptive Opportunistic Criminals , 2015, AAMAS.

[36]  John K. Maingi,et al.  Spatiotemporal patterns of elephant poaching in south-eastern Kenya , 2012, Wildlife Research.

[37]  R. McKelvey,et al.  Quantal Response Equilibria for Normal Form Games , 1995 .

[38]  Nigel Leader-Williams,et al.  Illegal exploitation of black rhinoceros and elephant populations : patterns of decline, law enforcement and patrol effort in Luangwa Valley, Zambia , 1990 .

[39]  John Zeleznikow,et al.  Decision support systems for police: Lessons from the application of data mining techniques to “soft” forensic evidence , 2006, Artificial Intelligence and Law.

[40]  R. Swihart,et al.  Absent or undetected? Effects of non-detection of species occurrence on wildlife-habitat models , 2004 .

[41]  Tuomas Sandholm,et al.  Lossy stochastic game abstraction with bounds , 2012, EC '12.