Generalized additive modelling and zero inflated count data

By Simon C. Barry, Alan H. Welsh. 2002

313 citations recorded by the source corpus.

Subfields: Ecological Modeling, Environmental Engineering, Nature and Landscape Conservation

View Paper: https://www.semanticscholar.org/paper/99beb5f538d9b681be1686c6ae7a3cb93c4b364f

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 Environmental Engineering

In Nature and Landscape Conservation

Citations in the corpus

Can models of presence‐absence be used to scale abundance? Two case studies considering extremes in life history (2005, methodology): “More recently, Barry and Welsh (2002) suggest use of a zero-inflated count model.”
Modelling species distributions using regression quantiles (2007, unknown): “Also, in contrast to GLM and GAM requiring a two step modelling procedure (Barry & Welsh 2002), RQ models prove effective in dealing with zero-inflated count data, which are common in species distribution models of abundance data.”
Spatial models for distance sampling data: recent developments and future directions (2013, methodology): “Zero inflation in count data may be problematic, and two-stage approaches such as Barry & Welsh (2002) as well as more flexible response distributions made possible by Rigby&Stasinopoulos (2005) have yet to be exploited by those using distance sampling data.”
Mixed Effects Models and Extensions in Ecology with R (2009, methodology): “…derivatives (which can easily be determined and we leave this as an exercise for the reader), and then it is purely a matter of numerical optimisation, though we end up with a slightly different algorithm compared to Poisson GLM. Details can be found in Barry and Welsh (2002) or Hilbe (2007).”
Spatial fisheries ecology: Recent progress and future prospects (2008, result): “…carried by the zero-counts and produces unbiased predictions, it makes inference rather compli- cated as there are two sets of environmental effects to consider, those from the presence/absence and those from the abundance model — often with contradictory results (Barry and Welsh, 2002).”
Dynamic wildlife habitat models: Seasonal foods and mortality risk predict occupancy-abundance and habitat selection in grizzly bears (2010, methodology): “We used a zero-inflated count (Poisson) model (Barry and Welsh, 2002) to relate local occupancy-abundance with realised habitat.”
Winter distribution of blue crab Callinectes sapidus in Chesapeake Bay : application and cross-validation of a two-stage generalized additive model (2005, methodology): “The 2 stages may then be combined multiplicatively to yield an overall prediction (Barry & Welsh 2002).”
Waterbird Populations and Pressures in the Baltic Sea (2011, methodology): “We used a two-step GAM approach (in literature also called a hurdle model or a delta model) to deal with zero inflation (an excess of zeros) in our data set (Stefánsson 1996, Barry & Welsh 2002, Potts and Elith 2006).”
Using delta generalized additive models to produce distribution maps for spatially explicit ecosystem models (2014, methodology): “Therefore, the delta GAM approach offers a perspective on patial patterns of abundance different from that offered by the egative binomial GAM approach, by identifying factors influencng the presence of species and environmental conditions favoring igh abundances (Barry and Welsh, 2002; Koubbi et al”
Modelling distribution and abundance with presence‐only data (2006, background): “Alternatively, it may be possible to modify zero-inflated Poisson or negative binomial (ZIP or ZINB) regression models (Welsh et al. 1996; Barry & Welsh 2002; Dirnböck & Dullinger 2004; Nielsen et al. 2005) to model abundance given availability, where available locations are assigned a value of…”
Predicting reptile distributions at the mesoscale: relation to climate and topography (2003, background): “Similarly, species’ abundance would be best simulated by a zero-inflated Poisson distribution (Barry & Welsh, 2002).”