gbm.auto

gbm.auto implements Boosted Regression Trees (BRT) to model and map species abundance from spatial environmental predictor data.

Key Features:

• Automation of Boosted Regression Trees (BRT): Automates key steps in BRT model fitting and configuration, including parameter optimization.
• Spatial modelling capabilities: Handles spatial data and environmental predictor variables to generate abundance maps.
• Ensemble decision-tree modelling: Uses BRT as an ensemble of multiple decision trees to improve predictive accuracy.
• Large dataset handling: Described as capable of handling large datasets efficiently during model fitting.
• Reproducibility and consistency: Promotes reproducible and consistent model fitting across studies.

Scientific Applications:

• Ecological research: Predicts species distribution patterns and responses to environmental change, habitat fragmentation, and climate change from abundance data.
• Conservation biology: Identifies critical habitats and assesses impacts of human activities on biodiversity to inform conservation strategies.
• Environmental monitoring: Maps species abundance to support detection of ecosystem shifts over time.

Methodology:

Implements Boosted Regression Trees (BRT) as an ensemble of decision trees, automates fitting to spatial data, performs parameter optimization, handles large datasets, and aims to improve predictive accuracy and reproducibility.