Boosted Beta Regression

Schmid, Matthias; Wickler, Florian; Maloney, Kelly O.; Mitchell, Richard M.; Fenske, Nora; Mayr, Andreas

doi:10.1371/journal.pone.0061623

Cited by 59 publications

(55 citation statements)

References 49 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…As sex‐specific topologies are tested independently, scaling was achieved using the maximum norm, consisting in dividing centrality scores by the maximum score of the centrality vector (Ruhnau, ). Fitting ordinary least‐squares regression to bounded and skewed response variable shows, however, the risk of violation of the linearity, normality and homoscedasticity assumptions (Cribari‐Neto & Zeileis, ; Schmid et al., ). Although transformation of the response distribution (e.g., logit or arcsine functions) is commonly used for complying with model assumptions, the resulting estimates might have biased predictive capacity and may be difficult to interpret (Cribari‐Neto & Zeileis, ; Warton & Hui, ).…”

Section: Methodsmentioning

confidence: 99%

“…Considering the intrinsic nature of the response variable distributions, we conducted beta regressions with the Betareg package (Cribari‐Neto & Zeileis, ) to investigate the relationship between node centrality of each sex‐specific graph (i.e., dependent variable) over the proportion of immigrants and the snow depth (i.e., independent variables). Because the model assumes that values of the response variable ( y ) lie strictly on the unit interval (0 < y < 1), we fitted limit boundaries of y by ( y ·( n − 1) + 0.5)/ n , where n corresponds to the sample size (Schmid et al., ). We also controlled for potential random variation in sampling intensities by adding the number of individual males and females sampled within each node as regressor of their respective sex‐specific model.…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Sex‐specific graphs: Relating group‐specific topology to demographic and landscape data

et al. 2017

View full text Add to dashboard Cite

Sex-specific genetic structure is a commonly observed pattern among vertebrate species. Facing differential selective pressures, individuals may adopt sex-specific life history traits that ultimately shape genetic variation among populations. Although differential dispersal dynamics are commonly detected in the literature, few studies have used genetic structure to investigate sex-specific functional connectivity. The recent use of graph theoretic approaches in landscape genetics has demonstrated network capacities to describe complex system behaviours where network topology represents genetic interaction among subunits. Here, we partition the overall genetic structure into sex-specific graphs, revealing different male and female dispersal dynamics of a fisher (Pekania [Martes] pennanti) metapopulation in southern Ontario. Our analyses based on network topologies supported the hypothesis of male-biased dispersal. Furthermore, we demonstrated that the effect of the landscape, identified at the population level, could be partitioned among sex-specific strata. We found that female connectivity was negatively correlated with snow depth, whereas connectivity among males was not. Our findings underscore the potential of conducting sex-specific analysis by identifying landscape elements or configuration that differentially promotes or impedes functional connectivity between sexes, revealing processes that may otherwise remain cryptic. We propose that the sex-specific graph approach would be applicable to other vagile species where differential sex-specific processes are expected to occur.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Sex‐specific graphs: Relating group‐specific topology to demographic and landscape data

et al. 2017

View full text Add to dashboard Cite

show abstract

“…The parametrization of the negative binomial distribution, the log-logistic distribution and the t distribution in boosted GAMLSS models is given in Mayr et al (2012a). The derivation of boosted beta regression, another special case of GAMLSS, can be found in Schmid, Wickler, Maloney, Mitchell, Fenske, and Mayr (2013). In our case study we will use the default GaussianLSS() family to model childhood malnutrition in India.…”

Section: Distributionsmentioning

confidence: 99%

gamboostLSS: An R Package for Model Building and Variable Selection in the GAMLSS Framework

Hofner¹,

Mayr²,

Schmid³

2016

J. Stat. Soft.

Self Cite

View full text Add to dashboard Cite

Generalized additive models for location, scale and shape are a flexible class of regression models that allow to model multiple parameters of a distribution function, such as the mean and the standard deviation, simultaneously. With the R package gamboostLSS, we provide a boosting method to fit these models. Variable selection and model choice are naturally available within this regularized regression framework. To introduce and illustrate the R package gamboostLSS and its infrastructure, we use a data set on stunted growth in India. In addition to the specification and application of the model itself, we present a variety of convenience functions, including methods for tuning parameter selection, prediction and visualization of results. The package gamboostLSS is available from the Comprehensive R Archive Network (CRAN) at https://CRAN.R-project.org/package=gamboostLSS.

show abstract

“…Another class of GLM that can model proportions or probabilities directly is beta regression (Ferrari & Cribari‐Neto, ). This approach is less commonly employed in ecology, although methods such as boosted beta regression are gaining popularity in the field (Schmid et al, ).…”

Section: Sdmsmentioning

confidence: 99%

Modelling imperfect presence data obtained by citizen science

et al. 2017

View full text Add to dashboard Cite

There is growing awareness about the potential benefit of harnessing citizen science for research, particularly in the biological and environmental sciences. Data quality is a major constraint in the use of citizen‐science data, in particular, imperfect observations. In this paper, we fit species distribution models to presence‐only data (presences and counts, with no absences observed) by exploiting the uncertainty in reported presences, instead of generating pseudo‐absences as is common in previous presence‐only studies. This approach allowed us to extend the suite of models to include those commonly fit to presence/absence and abundance data. We fit several models to a case study data set of jaguar encounters reported by citizens in the Peruvian Amazon. The true species distribution for the case study data is unknown, and thus we also undertake an extensive simulation study to evaluate model performance. We analyze the sources of error by studying the bias and variance of the models and discuss the predictive performance of each model and its ability to recover the true species distribution. The simulation study shows that, although several approaches are capable of recovering the species distribution, the choice of a modelling approach is a complex one and depends on factors such as inferential aim, model complexity, sample size, and computational resources. This study also addresses some issues in dealing with compound‐imperfect observations arising from citizen‐science data, and we discuss further steps needed in this research area.

show abstract

Boosted Beta Regression

Cited by 59 publications

References 49 publications

Sex‐specific graphs: Relating group‐specific topology to demographic and landscape data

Sex‐specific graphs: Relating group‐specific topology to demographic and landscape data

gamboostLSS: An R Package for Model Building and Variable Selection in the GAMLSS Framework

Modelling imperfect presence data obtained by citizen science

Contact Info

Product

Resources

About