Evaluating the predictive performance of habitat models developed using logistic regression

By J. Pearce, Simon Ferrier. 2000

2,040 citations recorded by the source corpus.

Subfields: Ecological Modeling, Ecology, Nature and Landscape Conservation

View Paper: https://www.semanticscholar.org/paper/078030d08092166ec6797d3e348fb7c579017e84

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Ecological Modeling

In Ecology

In Nature and Landscape Conservation

Citations in the corpus

Dealing with uncertain absences in habitat modelling: a case study of a rare ground‐dwelling parrot (2007, background): “The area under the curve (AUC) provides a measure of discrimination ability, and varies from 0.5 for a model with discrimination ability no better than random, to 1.0 for a model with perfect discriminatory ability (Pearce & Ferrier, 2000).”
The area under the precision‐recall curve as a performance metric for rare binary events (2019, methodology): “As a measure of (lack of) calibration, we calculated the absolute difference between the slope of Miller's calibration plot and one, as a perfectly calibrated model would have a slope of one (Miller, Hui, & Tierney, 1991; Pearce & Ferrier, 2000).”
Threshold criteria for conversion of probability of species presence to either–or presence–absence (2007, background): “…map into a categorical one involves adopting a threshold probability indicative of species presence (Fielding and Bell, 1997) which will determine model output, as it will condition the cases assigned to each category (Fielding and Bell, 1997; Manel et al., 1999b; Pearce and Ferrier, 2000).”
Why do some species have geographically varying responses to fire history (2014, methodology): “We used 7-fold cross validation for within-region predictive capacity (Pearce and Ferrier 2000).”
Modelling habitat and spatial distribution of an endangered longhorn beetle - A case study for saproxylic insect conservation (2007, methodology): “ROC-curves are a useful evaluation tool for predictive species distribution models (see Fielding and Bell, 1997; Pearce and Ferrier, 2000; Reineking and Schröder, 2006).”
Hollow futures? Tree decline, lag effects and hollow‐dependent species (2013, methodology): “The area under the receiver operating characteristic curve (AUC) for this model was 0.93, indicating the model had ‘excellent’ discriminating ability (Pearce & Ferrier, 2000).”
Connectivity, Probabilities and Persistence: Comparing Reserve Selection Strategies (2006, methodology): “An advantage is that probabilities can be modelled and estimated using a standard statistical method (e.g. logistic regression, which is a standard technique in this context; Austin et al. 1984; Ter Braak and Looman 1986; Pearce and Ferrier 2000).”
Regional variation in habitat–occupancy thresholds: a warning for conservation planning (2008, methodology): “AUCs for the most parsimonious models were slightly less than 0·7 (see Supplementary material, Table S1), indicating poor to reasonable discrimination ability (Pearce & Ferrier 2000).”
Patterns and uncertainties of species' range shifts under climate change (2004, methodology): “%) and tested on the remaining data sets with both the receiver–operating characteristic (ROC) curve and the k statistic (Pearce & Ferrier, 2000; Thuiller et al., 2003).”
Novel methods improve prediction of species' distributions from occurrence data (2006, neutral): “While other studies indicate that small sample size negatively influences modelling success (Pearce and Ferrier 2000a, Harrell 2001, Stockwell and Peterson 2002, Kadmon et al. 2003), this was not evident here, perhaps because in general we had adequate localities for modelling (mean number of…”
Climate Change and Biodiversity: Some Considerations in Forecasting Shifts in Species' Potential Distributions (2005, background): “…and chi-square tests (Anderson et al. 2003; Elith and Burgman 2002; Fleishman et al. 2003; Oreskes et al. 1994); detailed reviews of statistical methods can be found elsewhere (Fielding and Bell 1997; Pearce and Ferrier 2000; Pearce et al. 2001), but a couple of comments will be provided here.”
Predicting carnivore occurrence with noninvasive surveys and occupancy modeling (2011, background): “AUC values range from 0.5 (i.e., no better than a null model) to 1.0 (i.e., perfect accuracy; Fielding and Bell 1997; Pearce and Ferrier 2000).”
Temperature and Flow Effects on Migration Timing of Chinook Salmon Smolts (2009, methodology): “We used the receiver operating characteristic (ROC) curve and withheld data to evaluate the proportion of correctly and incorrectly classified migration events ( Pearce and Ferrier 2000 ).”
A comparative evaluation of presence‐only methods for modelling species distribution (2007, methodology): “Pearce & Ferrier (2000) suggested that algorithms used to fit species distribution models can be ranked according to their ‘function complexity’.”