Premium
Evaluating Lost Person Behavior Models
Author(s) -
Sava Elena,
Twardy Charles,
Koester Robert,
Sonwalkar Mukul
Publication year - 2016
Publication title -
transactions in gis
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.721
H-Index - 63
eISSN - 1467-9671
pISSN - 1361-1682
DOI - 10.1111/tgis.12143
Subject(s) - watershed , terrain , metric (unit) , statistics , wilderness , euclidean distance , joint probability distribution , probability distribution , computer science , geography , mathematics , artificial intelligence , cartography , operations management , engineering , machine learning , biology , ecology
US wilderness search and rescue consumes thousands of person‐hours and millions of dollars annually. Timeliness is critical: the probability of success decreases substantially after 24 hours. Although over 90% of searches are quickly resolved by standard “reflex” tasks, the remainder require and reward intensive planning. Planning begins with a probability map showing where the lost person is likely to be found. The M ap S core project described here provides a way to evaluate probability maps using actual historical searches. In this work we generated probability maps the E uclidean distance tables in ( K oester [Koester R J, 2008]), and using D oke's ([Doke J, 2012]) watershed model. W atershed boundaries follow high terrain and may better reflect actual barriers to travel. We also created a third model using the joint distribution using E uclidean and watershed features. On a metric where random maps score 0 and perfect maps score 1, the Euclidean distance model scored 0.78 (95% CI : 0.74–0.82, on 376 cases). The simple watershed model by itself was clearly inferior at 0.61, but the Combined model was slightly better at 0.81 (95% CI : 0.77–0.84).