A working guide to boosted regression trees
2008/04/08 by Jane Elith, J. Elith, J. R. Leathwick +3 · 169 citations
Environmental Science · Computer Science · #Fish Ecology and Management Studies #Species Distribution and Climate Change #Data Analysis with R
paper · pdf · doi:10.1111/j.1365-2656.2008.01390.x
Abstract
1. Ecologists use statistical models for both explanation and prediction, and need techniques that are flexible enough to express typical features of their data, such as nonlinearities and interactions. 2. This study provides a working guide to boosted regression trees (BRT), an ensemble method for fitting statistical models that differs fundamentally from conventional techniques that aim to fit a single parsimonious model. Boosted regression trees combine the strengths of two algorithms: regression trees (models that relate a response to their predictors by recursive binary splits) and boosting (an adaptive method for combining many simple models to give improved predictive performance). The final BRT model can be understood as an additive regression model in which individual terms are simple trees, fitted in a forward, stagewise fashion. 3. Boosted regression trees incorporate important advantages of tree-based methods, handling different types of predictor variables and accommodating missing data. They have no need for prior data transformation or elimination of outliers, can fit complex nonlinear relationships, and automatically handle interaction effects between predictors. Fitting multiple trees in BRT overcomes the biggest drawback of single tree models: their relatively poor predictive performance. Although BRT models are complex, they can be summarized in ways that give powerful ecological insight, and their predictive performance is superior to most traditional modelling methods. 4. The unique features of BRT raise a number of practical issues in model fitting. We demonstrate the practicalities and advantages of using BRT through a distributional analysis of the short-finned eel (Anguilla australis Richardson), a native freshwater fish of New Zealand. We use a data set of over 13 000 sites to illustrate effects of several settings, and then fit and interpret a model using a subset of the data. We provide code and a tutorial to enable the wider use of BRT by ecologists.
Cited by
- Why aren't rabbits and hares larger?
- Microbial functional guilds and genes are key to explaining soil nutrient cycling alongside soil and plant variables
- Fundamental Interaction Niches: Towards a Functional Understanding of Ecological Networks' Resilience
- Towards a predictive framework for biocrust mediation of plant performance: A meta‐analysis
- Biogeography and evolutionary patterns of temperate deciduous forests in the Northern Hemisphere
- Integrating ecosystem metabolism and consumer allochthony reveals nonlinear drivers in lake organic matter processing
- Changes in community composition and functional diversity of European bats under climate change
- Automated machine learning for pollen-based reconstructions: Evaluation with modern pollen dataset from China and application to Oligocene records
- Introduction to deep learning methods for multi‐species predictions
- Perched on a razor's edge: Woodlark <i>Lullula arborea</i> preference for vanishing land-uses in a traditional sub-Pannonian landscape
- Does the abiotic environment influence the distribution of flower and fruit colors?
- Linking N2O emissions and nosZ gene abundance: A meta-analysis of organic carbon amendments in agricultural soils
- Investigating Morphospace Occupation in Multi-Scale Ecological and Evolutionary Data Using Regression Tree: Case Studies and Perspectives
- Berries and bullets: influence of food and mortality risk on grizzly bears in British Columbia
- Limited evidence for the impact of climate change on the recent range expansion of the malaria vector <i>Anopheles stephensi</i> in the Horn of Africa
- Restoration cannot be scaled up globally to save reefs from loss and degradation
- Correcting environmental sampling bias improves transferability of species distribution models
- Climate change effects on marine protected areas: Projected decline of benthic species in the North Sea
- A productivity bottleneck in the Baltic herring (Clupea harengus membras): Early life-history processes and recruitment variability
- Acoustic telemetry suggests the lesser spotted dogfish Scyliorhinus canicula stays and uses habitats within a French offshore wind farm
- Future climate change is predicted to shift long-term persistence zones in the cold-temperate kelp Laminaria hyperborea
- Uncertainty in blacklisting potential Pacific plant invaders using species distribution models
- Improving species distribution models by optimising background points: Impacts on current and future climate projections
- Within‐Night Variation in Predictor Importance Highlights Dynamic Nature of Bird Migration
- Monitoring of hive weight changes in various landscapes
- Machine Learning Applications for Fisheries—At Scales from Genomics to Ecosystems
- Predicted occurrence and abundance habitat suitability of invasive plants in the contiguous United States: updates for the INHABIT web tool
- Navigating Future Waters: The Resilience of the Atlantic Bluefin Tuna Under Climate Change
- Exploring the determinants of the 2023 Quebec, Canada, mega-wildfire burn severity
- Global intraspecific diversity of marine forests of brown macroalgae predicted by past climate conditions
- Comparative Evaluation of Tree-Based Ensemble Algorithms for Short-Term\n Travel Time Prediction
- Projecting bird assemblage responses to climate-driven changes in managed boreal forest landscapes of Québec
- Weighted batch means estimators in Markov chain Monte Carlo
- Methodological Techniques Used in Machine Learning to Support Individualized Drug Dosing Regimens Based on Pharmacokinetic Data: A Scoping Review
- Predicting potential distributions of freshwater invasive species in novel environments
- Predicting Pediatric Surgical Durations
- Cetaceans and sea turtles in the northern region of the Mediterranean Cetacean Migration Corridor: abundance and multi-model habitat suitability analysis
- Where is positional uncertainty a problem for species distribution modelling?
- A standard protocol for reporting species distribution models
- An identified agronomic interpretation for potassium permanganate oxidizable carbon
- Regional and global elevational patterns of microbial species richness and evenness
- Urbanization is associated with unique community simplification among birds in a neotropical landscape
- Temporal and spatial transferability in telemetry-based dynamic species distribution models: The effects of algorithms and pseudo-absence techniques
- GIS-based groundwater potential mapping using boosted regression tree, classification and regression tree, and random forest machine learning models in Iran
- High-resolution ocean sea ice model outputs to predict species distribution: case study of crabeater seal Lobodon carcinophagus in East Antarctica
- Leveraging long‐term data to improve biodiversity monitoring with species distribution models
- Submarine canyons and oceanographic processes shape suitable habitats of the deep-sea shrimp Aristaeopsis edwardsiana (Johnson, 1868) in the Southwestern Atlantic
- Gradient Boosted Mixed Models: Flexible Estimation of Mean and Variance Components for Clustered Data
- Machine Intelligent Techniques for Ramp Event Prediction in Offshore and Onshore Wind Farms
- Potential Spatial Mismatches Between Marine Predators and Their Prey in the Southern Hemisphere in Response to Climate Change
- Ecology with artificial intelligence and machine learning in Asia: A historical perspective and emerging trends
- Machine learning models reveal the spatial distribution and key environmental drivers of Coilia nasus in East China Sea's coastal waters
- The Need to Manage Prey Fish Biomass to Support Shark Conservation
- Effectiveness of territorial protection categories in conserving an endemic island bird during population fluctuation
- Relative role of operational patterns and environmental conditions on cookie cutter shark damage in the Hawai‘i longline fishery
- 38 years of declining and stable chlorophyll-a trends in northern US rivers
- Predicting Road‐Crossing Passability for River Connectivity Analysis
- Mismatch between climate-based bioenergetics model of fish growth and long-term and regional-scale empirical data
- Drivers of phenological transitions in the seedling life stage
- Microbial functional guilds and genes are key to explaining soil nutrient cycling alongside soil and plant variables
- Coupling Diversity Indices with Boosted Regression Trees Predicts Species Abundance and Spatial Community Patterns in Dry Afromontane Forests
- Distribution and vegetation representation of pollen assemblages from surface sediments of Nam Co, a large alpine lake in the central Tibetan Plateau
- Atmosphere, Vegetation, and Soil Water Coupling Determined by Stomatal Regulation of Transpiration
- Optimizing deep neural networks for estimating soil water retention curves: A comparison of metaheuristic and numerical algorithms
- Impact of glyphosate on soil microbial biomass and respiration: A meta-analysis
- Seabird‐vessel interactions in industrial fisheries of Northwest Africa: Implications for international bycatch management
- geoSABINA: a unified plant ecology database for Spain
- Contrasting effects of urbanization on vegetation between the Global South and Global North
- An efficient protocol for the global sensitivity analysis of stochastic ecological models
- Unveiling the patchwork: landscape arrangement drives moose winter utilization of forests
- Adaptive capacity within tropical marine protected areas – Differences between men- and women-headed households
- Modelling soil moisture in a high‐latitude landscape using LiDAR and soil data
- Gene drives for vertebrate pest control: Realistic spatial modelling of eradication probabilities and times for island mouse populations
- Increasing Behavioral Flexibility? An Integrative Macro-Scale Approach to Understanding the Middle Stone Age of Southern Africa
- MFAI: A Scalable Bayesian Matrix Factorization Approach to Leveraging Auxiliary Information
- Intrinsic traits of woodland caribou Rangifer tarandus caribou calves depredated by black bears Ursus americanus and coyotes Canis latrans
- Employees recruitment: A prescriptive analytics approach via machine learning and mathematical programming
- BON in a Box: An Open and Collaborative Platform for Biodiversity Monitoring, Indicator Calculation, and Reporting
- Can seed banking assist in conserving the highly endemic New Zealand indigenous flora?
- The effect of trait choice on hybrid species distribution model projections under climate change
- Impacts of nonnative fishes on local contributions to beta diversity in United States freshwater fishes
- When Good Fit Goes Bad: Identifying and Minimising Overfitting in Ecological Niche Models
- An unmanned surface vehicle for acoustic telemetry surveys of coastal fishes: range testing, real world performance, and comparison with a stationary tracking array
- Explainable machine learning improves interpretability in the predictive modeling of biological stream conditions in the Chesapeake Bay Watershed, USA
- Batch size selection for variance estimators in MCMC
- Improving propensity score weighting using machine learning
- Reshuffling of Azorean Coastal Marine Biodiversity Amid Climate Change
- Extinction risk of the endemic soft coral Phyllogorgia dilatata (Esper, 1806) in a remote island of the Brazilian coast
- Scattered woody vegetation promotes European brown hare population
- Machine Learning Based Routing Congestion Prediction in FPGA High-Level Synthesis
- Historical perspective of artisanal encircling gillnet use at the Brazilian coast: Changes in fishing behaviour is mirrored by dwindling stocks
- Spatial variability in herbaceous plant phenology is mostly explained by variability in temperature but also by photoperiod and functional traits
- Boosted Regression Trees Machine Learning Method Improves the brGDGT‐Based Climate Reconstruction in Drylands
- Mapping the geography of livestock systems in South America’s dry diagonal
- Similar drivers but distinct patterns of woody and herbaceous alien plant invasion
- Key drivers of at‐vessel mortality in demersal sharks
- The Vulnerability of Overwintering Insects to Loss of the Subnivium
- Functional Traits and Phylogeny Explain Plant Phenological Variations Among Native and Non‐Native Species in Botanical Gardens
- Refined gap analysis for biodiversity conservation under climate change
- Drivers of amphibian species richness in European ponds
- Genetic Monitoring of a Lethal Control Programme for Wild Canids With Complex Mating Strategies
- Herbarium collections remain essential in the age of community science
- Machine learning meets ecology: XGBoost-based prediction of endangered species habitats using multi-source environmental data
- Seasonal trophic controls drive population variability in a foundational marine copepod
- Environmental and spatial correlates of hydrologic alteration in a large Mediterranean river catchment
- North–south Holocene seasonal contrast on the Italian peninsula: Clarification of the latitudinal transition zone based on the multi-proxy study of Lago Grande di Monticchio
- Photosynthetic optimum temperature plays a minor role in the increase of terrestrial carbon uptake from 2000 to 2019
- A conceptual framework for scaling up emergent predictions from mechanistic individual-based models
- What determines trophic niche breadth? A global analysis of freshwater fishes using isospaces
- Infestation symptoms as indicators of a sustained bark beetle outbreak in conserved and managed Norway spruce forests in south-eastern Finland
- Paludiculture maintains peat formation potential in rewetted temperate fens
- The thermal structure of small and shallow Arctic Fennoscandian lakes
- Environmental specialization decouples geographic range and habitat occupancy in Ragala ucuquirana-branca (Sapotaceae) across the Amazon Basin
- Soil microbiome and ecological traits depending on fertilization and climate in organic input agricultural ecosystem
- ‘specleanr': an R package for automated flagging of environmental outliers in ecological data for modeling workflows
- Multi-year migrations of four coastal shark species in the southeastern USA: trends in timing and temperature
- Uncertainty in blacklisting potential Pacific plant invaders using species distribution models
- Long‐term dynamics of coastal dune landscapes and habitat diversity: Insights from a quarter century of resurveys in Castelporziano Presidential Estate
- Re-scale boosting for regression and classification
- Harnessing artificial intelligence to fill global shortfalls in biodiversity knowledge
- Projected impacts of future climate change on the aboveground biomass of seagrasses at global scale
- Predictive Modeling of Juvenile Smalltooth Sawfish Habitats: Challenges and Opportunities for Conservation
- A new approach to derive productivity of tropical forests using radar remote sensing measurements
- Climate change reshapes species-specific connectivity among marine protected areas in the Yangtze River estuary and adjacent waters
- Climate-driven habitat suitability of Tinospora cordifolia under changing climate conditions
- Machine learning-based ensemble species distribution models to guide monitoring and survey design for offshore wind
- Breeding habitat suitability of the endemic Tristram’s Warbler Curruca deserticola in North Africa: a micro- and macro-scale approach
- Felid distributions in the Southern Yungas: enhancing species distribution models with field data
- Interspecific Habitat Suitability of Four Southeast Asian Spiny Climbing Palms (Korthalsia) Through Species Distribution Modeling
- A framework for climate-resilient forest planning and restoration: advances linking species distribution modelling, genetic adaptation and future climate scenarios
- Identifying suitable habitats under climate change for non-targeted demersal fish in the Mediterranean Sea
- Determinants of Terrestrial and Limnic Species Richness in Germany
- Insurance Premium Prediction via Gradient Tree-Boosted Tweedie Compound Poisson Models
- A XGBoost risk model via feature selection and Bayesian hyper-parameter optimization
- DART: Dropouts meet Multiple Additive Regression Trees
- Alternative plant designs: consequences for community assembly and ecosystem functioning
- Interpreting Black-Box Models: A Review on Explainable Artificial Intelligence
- Testing whether ensemble modelling is advantageous for maximising predictive performance of species distribution models
- Predicting Novel Tick Vectors of Zoonotic Disease
- Climate change exposure of waterbird species in the African-Eurasian flyways
- Changes in soil organic carbon in native forests of Argentina related to land use change and environmental factors
- U.S. National VS30 Models and Maps Informed by Remote Sensing and Machine Learning
- microRNA‐based predictor for diagnosis of frontotemporal dementia
- PaloBoost: An Overfitting-robust TreeBoost with Out-of-Bag Sample Regularization Techniques
- Improving propensity score weighting using machine learning. [europepmc]
- The dominant Anopheles vectors of human malaria in the Americas: occurrence data, distribution maps and bionomic précis. [europepmc]
- The dominant Anopheles vectors of human malaria in Africa, Europe and the Middle East: occurrence data, distribution maps and bionomic précis. [europepmc]
- The dominant Anopheles vectors of human malaria in the Asia-Pacific region: occurrence data, distribution maps and bionomic précis. [europepmc]
- A global map of dominant malaria vectors. [europepmc]
- Predicting the risk of avian influenza A H7N9 infection in live-poultry markets across Asia. [europepmc]
- Global distribution maps of the leishmaniases. [europepmc]
- Mapping the zoonotic niche of Ebola virus disease in Africa. [europepmc]
- Childhood Maltreatment, Depression, and Suicidal Ideation: Critical Importance of Parental and Peer Emotional Abuse during Developmental Sensitive Periods in Males and Females. [europepmc]
- The global distribution of the arbovirus vectors Aedes aegypti and Ae. albopictus. [europepmc]
- The ecology of microscopic life in household dust. [europepmc]
- Mapping global environmental suitability for Zika virus. [europepmc]
- Geographical distributions of African malaria vector sibling species and evidence for insecticide resistance. [europepmc]
- Estimating the High-Arsenic Domestic-Well Population in the Conterminous United States. [europepmc]
- Global hotspots and correlates of emerging zoonotic diseases. [europepmc]
- The Impact of Acute Organ Dysfunction on Long-Term Survival in Sepsis. [europepmc]
- Drinking Water Nitrate and Human Health: An Updated Review. [europepmc]
- Wildfires and climate change push low-elevation forests across a critical climate threshold for tree regeneration. [europepmc]
- The current and future global distribution and population at risk of dengue. [europepmc]
- Microdiversity ensures the maintenance of functional microbial communities under changing environmental conditions. [europepmc]
- Artificial Intelligence and Machine Learning in Pathology: The Present Landscape of Supervised Methods. [europepmc]
- A data-driven approach to predicting diabetes and cardiovascular disease with machine learning. [europepmc]
- Global meta-analysis shows pervasive phosphorus limitation of aboveground plant production in natural terrestrial ecosystems. [europepmc]
- A new malaria vector in Africa: Predicting the expansion range of Anopheles stephensi and identifying the urban populations at risk. [europepmc]
- Mapping ticks and tick-borne pathogens in China. [europepmc]
Related