CFE 2019: Start Registration
View Submission - CMStatistics
B1928
Title: Predicting shape of dives of Southern elephant seals using functional regression tree models Authors:  Morgan Godard - MIO-Mediterranean Institute of Oceanography (France) [presenting]
Claude Mante - Aix-Marseille University (France)
Christophe Guinet - CEBC (France)
David Nerini - Mediterranean Institute of Oceanology (France)
Abstract: In recent years, the study of animal movements in the ocean has been revolutionized with massive use of miniature measuring devices providing access to complex behavioral data and associated environmental data sampled at very high frequency. These data are called functional data. The objectives are to highlight the relationships between dives of Southern elephant seals, Mirounga leonina, and the physical environment in which elephant seals operate. Starting from a huge data set of elephant seal dives, we first show how to construct functional dive profiles from point-wise samples. We then propose a generalized regression tree method where the predictive variable is a function. Regression tree models are built to predict the shape of dives using discrete environmental variables (i.e. temperature at a depth of 250m) and environmental profiles (i.e. temperature, salinity) as predictors. The connection between shapes of dives and shapes of environmental profiles will be highlighted. Tree capabilities for predictor selection and the choice of splitting criterion will be discussed.