Species Distribution Modeling via Spatial Bagging of Multiple Conditional Random Fields
Satellite tracking technologies enable scientists to collect data of animal migrations and species habitats on a large scale. Modeling distributions of wild animals is of considerable use. It helps researchers to understand important ecological phenomena such as the spread of bird flu and climate changes. Species distribution modeling has been studied for a long time, however, most existing work provide solutions in a point-wise manner, ignoring the relevance between adjacent habitats, which may reflect an important dependency between nearby places. In this paper, we take the relevance into consideration, and then propose a novel method to model species habitats and predict possible distribution of wild animals by applying the Spatial Bagging of Multiple Conditional Random Fields(SBMCRFs) on remote-sensing data. To access the usability of our method, several experiments are implemented on a real world dataset of migratory birds from Qinghai Lake Reserve. The experiment results show that SBMCRFs outperforms the baselines significantly, and the relevance between nearby places is demonstrated to be an important factor in species distribution modeling.
KeywordsSpecies distribution modeling Conditional random fields Ensemble methods
This work is partly supported by the Natural Science Foundation of China (NSFC) under Grant No. 41371386 and 91224006.
- Busby, J.: BIOCLIM a bioclimate analysis and prediction system. Plant Prot. Q., 6 (1991)Google Scholar
- Etherington, T.R., Ward, A.I., Smith, G.C., Pietravalle, S., Wilson, G.J.: Using the Mahalanobis distance statistic with unplanned presence only survey data for biogeographical models of species distribution and abundance: a case study of badger setts. J. Biogeogr. 36(5), 845–853 (2009)CrossRefGoogle Scholar
- Zheng, V.W., Zheng, Y., Xie, X., Yang, Q. Collaborative location, activity recommendations with GPS history data. In: Proceedings of the 19th International Conference on World Wide Web, pp. 1029–1038. ACM (2010)Google Scholar
- Li, Z., Han, J., Ji, M., Tang, L.A., Yu, Y., Ding, B., Kays, R.: Movemine: mining moving object data for discovery of animal movement patterns. ACM Trans. Intell. Syst. Technol. (TIST) 2(4), 37 (2011)Google Scholar
- Tang, M.J., Zhou, Y.C., Cui, P., Wang, W., Li, J., Zhang, H., Hou, Y.S., Yan, B.P.: Discovery of migration habitats and routes of wild bird species by clustering and association analysis. In: Huang, R., Yang, Q., Pei, J., Gama, J., Meng, X., Li, X. (eds.) ADMA 2009. LNCS, vol. 5678, pp. 288–301. Springer, Heidelberg (2009)CrossRefGoogle Scholar
- Tang, M.J., Wang, W., Jiang, Y., Zhou, Y., Li, J., Cui, P., Liu, Y., Yan, B.: Birds bring flues? mining frequent and high weighted cliques from birds migration networks. In: Kitagawa, H., Ishikawa, Y., Li, Q., Watanabe, C. (eds.) DASFAA 2010. LNCS, vol. 5982, pp. 359–369. Springer, Heidelberg (2010)CrossRefGoogle Scholar
- Pearson, R.G.: Species distribution modeling for conservation educators and practitioners. Lessons in Conservation (LinC) Developing the capacity to sustain the earth’s diversity, 54 (2007)Google Scholar
- Caruana, R., Elhawary, M., Munson, A., Riedewald, M., Sorokina, D., Fink, D., Hochachka, W.M., Kelling, S.: Mining citizen science data to predict orevalence of wild bird species. In: Proceedings of the 12th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pp. 909–915. ACM, NewYork (2006)Google Scholar
- Lafferty, J., McCallum, A., Pereira, F.C.: Conditional random fields,: Probabilistic models for segmenting and labeling sequence data. In: Proceedings of the Eighteenth International Conference on Machine Learning, ICML 01, pp. 282–289. CA, USA, Morgan Kaufmann Publishers Inc, San Francisco (2001)Google Scholar