Support vector machines for predicting distribution of Sudden Oak Death in California

Guo, Qinghua; Kelly, Maggi; Graham, Catherine H.

doi:10.1016/j.ecolmodel.2004.07.012

Cited by 273 publications

(166 citation statements)

References 38 publications

Supporting

Mentioning

159

Contrasting

Unclassified

Order By: Relevance

“…SVM is a machine-learning method that belongs to a family of generalised linear classifiers. To estimate the potential distribution of a species subject to the environmental conditions, the eco-space (spanned by the en vironmental variables) is separated by a hyperplane into 2 target classes (Guo et al 2005): suitable and unsuitable environmental conditions. The optimality criterion used to find the separating hyperplane is maximised distance to the (nearest) training data points (large margin separation).…”

Section: Sdmsmentioning

confidence: 99%

Species distribution modelling of marine benthos: a North Sea case study

Reiss¹,

Cunze²,

König³

et al. 2011

Mar. Ecol. Prog. Ser.

186

127

View full text Add to dashboard Cite

Species distribution models (SDMs) were applied to predict the distribution of benthic species in the North Sea. An understanding of species distribution patterns is essential to gain insight into ecological processes in marine ecosystems and to guide ecosystem management strategies. Therefore, we compared 9 different SDM methods, including GLM, GBM, FDA, SVM, RF, MAXENT, BIOCLIM, GARP and MARS, by using 10 environmental variables to model the distribution of 20 marine benthic species. Most of the models showed good or very good performance in terms of predictive power and accuracy, with highest mean area under the curve (AUC) values of 0.845 and 0.840, obtained for the MAXENT and GBM models, respectively. The poorest performance was shown by the BIOCLIM model, which had a mean AUC of 0.708. Nevertheless, the mapped distribution patterns varied remarkably depending on the model used, with up to 32.5% differences in predictions between models. For species with a narrow distribution range, the models showed a better performance based on the AUC than for species with a broad distribution range, which can most likely be attributed to the restricted spatial scale and the model evaluation procedure. Of the environmental variables, bottom water temperature and depth had the greatest effect on the distribution of 14 benthic species, based on MAXENT results. We examine the potential utility of this strategy for predicting future distribution of benthic species in response to climate change.

show abstract

Section: Sdmsmentioning

confidence: 99%

Species distribution modelling of marine benthos: a North Sea case study

Reiss¹,

Cunze²,

König³

et al. 2011

Mar. Ecol. Prog. Ser.

186

127

View full text Add to dashboard Cite

show abstract

“…Considering this, we used a one-class classification method [52] instead of traditional two-class classifiers. One-class classification methods have been successfully used to extract a specific land cover class and detect related changes [53,55,56]. Only samples from the target class are used during the training process, and no information about the counterpart (outlier class) is required.…”

Section: Extraction Of Urban Areas Using Multiple Sensor Datamentioning

confidence: 99%

“…In this context, the problem of extracting urban extent from multi-sensor data possesses the features of the one-class classification problem, where only data from the target class (i.e., urban area in this study) are available and well sampled [51,52]. One-class classification methods have successfully been used in extracting specific land cover types and change types [53][54][55][56]. Among all one class classification methods adopted in the remote sensing field, one-class support vector machine (OCSVM) [54][55][56] and support vector data description (SVDD) [53] have been widely discussed.…”

Section: Introductionmentioning

confidence: 99%

Regional Urban Extent Extraction Using Multi-Sensor Data and One-Class Classification

Zhang

Cai

2015

Remote Sensing

View full text Add to dashboard Cite

Stable night-time light data from the Defense Meteorological Satellite Program (DMSP) Operational Line-scan System (OLS) provide a unique proxy for anthropogenic development. This paper presents a regional urban extent extraction method using a one-class classifier and combinations of DMSP/OLS stable night-time light (NTL) data, MODIS normalized difference vegetation index (NDVI) data, and land surface temperature (LST) data. We first analyzed how well MODIS NDVI and LST data quantify the properties of urban areas. Considering that urban area is the only class of interest, we applied the one-class support vector machine (OCSVM) to classify different combinations of the three datasets. We evaluated the effectiveness of the proposed method and compared with the locally optimized threshold method in regional urban extent mapping in China. The experimental results demonstrate that DMSP/OLS NTL data, MODIS NDVI and LST data provide different but complementary information sources to quantify the urban extent at a regional scale. The results also indicate that the OCSVM classification of the combination of all three datasets generally outperformed the locally optimized threshold method. The proposed method effectively and efficiently extracted the urban extent at a regional scale, and is applicable to other study areas.

show abstract

“…The past decade has seen an expansion in the types and sophistication of species distribution modeling approaches that can be used to forecast changes to species and communities based on locality data and baseline conditions (Graham et al 2004;Guisan and Thuiller 2005;Guo et al 2005;Elith et al 2006;Elith and Leathwick 2009). SDMs have many challenges when forecasting species responses under changing environments, including the potential absence of a species-environment equilibrium, the difficulty to account for dispersal limitations, biotic interactions, phenotypic plasticity and evolutionary changes, and the incidence of novel environments outside the range of conditions used to calibrate the models (Elith and Leathwick 2009).…”

Section: Model Assessment and Validationmentioning

confidence: 99%

Rescuing and Sharing Historical Vegetation Data for Ecological Analysis: The California Vegetation Type Mapping Project

Kelly¹

2016

Biodiv. Inf.

Self Cite

View full text Add to dashboard Cite

Abstract.-Research efforts that synthesize historical and contemporary ecological data with modeling approaches improve our understanding of the complex response of species, communities, and landscapes to changing biophysical conditions through time and in space. Historical ecological data are particularly important in this respect. There are remaining barriers that limit such data synthesis, and technological improvements that make multiple diverse datasets more readily available for integration and synthesis are needed. This paper presents one case study of the Wieslander Vegetation Type Mapping project in California and highlights the importance of rescuing, digitizing and sharing historical datasets. We review the varied ecological uses of the historical collection: the vegetation maps have been used to understand legacies of land use change and plan for the future; the plot data have been used to examine changes to chaparral and forest communities around the state and to predict community structure and shifts under a changing climate; the photographs have been used to understand changing vegetation structure; and the voucher specimens in combination with other specimen collections have been used for large scale distribution modeling efforts. The digitization and sharing of the data via the web has broadened the scope and scale of the types of analysis performed. Yet, additional research avenues can be pursued using multiple types of VTM data, and by linking VTM data with contemporary data. The digital VTM collection is an example of a data infrastructure that expands the potential of large scale research through the integration and synthesis of data drawn from numerous data sources; its journey from analog to digital is a cautionary tale of the importance of finding historical data, digitizing it with best practices, linking it with other datasets, and sharing it with the research community.

show abstract

Support vector machines for predicting distribution of Sudden Oak Death in California

Cited by 273 publications

References 38 publications

Species distribution modelling of marine benthos: a North Sea case study

Species distribution modelling of marine benthos: a North Sea case study

Regional Urban Extent Extraction Using Multi-Sensor Data and One-Class Classification

Rescuing and Sharing Historical Vegetation Data for Ecological Analysis: The California Vegetation Type Mapping Project

Contact Info

Product

Resources

About