Final Technical Report: - Southwest Fisheries Science Center - NOAA
Final Technical Report: - Southwest Fisheries Science Center - NOAA
Final Technical Report: - Southwest Fisheries Science Center - NOAA
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
When cetacean experts were shown geographic contour plots of the predictions from the<br />
competing simple and complex mgcv gam models for each species during the SWFSC Cetacean<br />
Experts’ Workshop, the simple models were overwhelmingly preferred to the complex models.<br />
The dominant criticisms of the complex models from the expert panel were twofold: the<br />
predictions from the complex models either 1) exhibited relatively small-scale details in<br />
population density that are unexplainable given existing knowledge of the dynamics of the<br />
ecosystem, or 2) were nearly identical to those from the simple model and, therefore, the extra<br />
model complexity was not necessary for capturing the spatial patterns.<br />
Overall conclusions to be made from this investigation into the behavior of mgcv gam<br />
(summarized in Tables 10 and 11) are as follows:<br />
4.2.2 Encounter Rate Models<br />
• The gam.method perf.magic produced the simple models with the greatest predictive<br />
performance. The best complex models were developed using outer (6 models), perf.outer (4<br />
models), and perf.magic (2 models).<br />
• Cubic regression splines were preferred for building simple encounter rate models,<br />
whereas the complex models were constructed using either cubic or thin plate regression splines.<br />
• To our surprise, the preferred simple models were split almost equally between those<br />
built using gamma = 1.0 (8 models) and 1.4 (6 models). The best complex models were<br />
generally constructed using gamma = 1.0.<br />
• The sum of absolute deviations of the observed-to-predicted ratios was smaller for the<br />
complex models in most instances, although this is to be expected because the predictions were<br />
based on the same data used to build the models for this exercise.<br />
4.2.3 Group Size Models<br />
• The gam.method magic produced the simple models with the greatest predictive<br />
performance. The best complex models were divided among gam.methods mgcv and magic.<br />
• The preferred simple models were constructed by thin plate regression splines, in general,<br />
whereas cubic regression splines were found in more of the preferred complex models.<br />
• The gamma parameter performed close to our expectations in the group size models, with<br />
the majority of simple models constructed using gamma = 1.4 and the majority of complex<br />
models using the default value of 1.0.<br />
• The trend in the sum of absolute deviations of the observed-to-predicted ratios was<br />
similar to that found for the encounter rate models, with simple models tending to have slightly<br />
larger values.<br />
49