Page 226 - Special Topic Session (STS) - Volume 2

STS486 R. Ayesha A. et al.
can help mitigate model overfit. In the empirical analysis, the BIC associated
with the penalized and unpenalized models were similar, but the penalized
model was much simpler and could be explained from an ecological
perspective.
Potential improvements to model selection consistency could be attained
by using an adjusted BIC. For example, Hui et al. (2015) suggest the extended
regularization information criteria (ERIC), a modification of the BIC, where the
model complexity term is also a function of ̃. Although we studied network
sizes motivated by ecological network sizes, extending the framework to the
high dimensional setting where K≫N likely only requires another modification
of the BIC, tailored specifically for the high dimensional case. Regardless,
regularized grouped DM regression is a useful tool for ecologists and can
provide insights into the functional traits driving species’ interactions.

References
1. Arnold, S. E., Savolainen, V. and Chittka, L. (2009). Flower colours along
an alpine altitude gradient, seen through the eyes of fly and bee
pollinators, Arthropod-Plant Interactions 3(1): 27-43.
2. Bascompte, J. and Jordano, P. (2007). Plant-Animal Mutualistic Networks:
The Architecture of Biodiversity, Annual Review of Ecology Evolution and
Systematics 38(1): 567-593.
3. Beck, A. and Teboulle, M. (2009). A fast iterative shrinkage-thresholding
algorithm for linear inverse problems, SIAM Journal on Imaging Sciences
2(1): 183-202.
4. Crea, C. (2017). Variable Selection for Grouped DM regression,
https://github.-com/ccrea/ VarSelect_DM_Regression.
5. Crea, C., Ali, R. A. and Rader, R. (2016). A new model for ecological
networks using species-level traits, Methods in Ecology and Evolution
7(2): 232-241.
6. Guimarães, P. and Lindrooth, R. C. (2007). Controlling for overdispersion
in grouped conditional logit models: A computationally simple
application of Dirichlet-multinomial regression, The Econometrics
Journal 10(2): 439-452.
7. Hui, F. K., Warton, D. I. and Foster, S. D. (2015). Tuning parameter
selection for the adaptive lasso using ERIC, Journal of the American
Statistical Association 110(509): 262-269.\
8. McFadden, D. (1974). Conditional Logit Analysis of Qualitative Choice
Behavior, in P. Zarembka (ed.), Frontiers in econometrics, Academic
Press, New York, pp. 105-142.
9. Montoya, J. M., Pimm, S. L. and Solé, R. V. (2006). Ecological networks
and their fragility, Nature 442(7100): 259-264.

215 | I S I W S C 2 0 1 9