MIAT: Modular R-wrappers for flexible implementation of MaxEnt distribution modelling

摘要: Abstract The maximum entropy (MaxEnt) method has gained widespread use for distribution modelling, mostly because of the practical simplicity offered by maxent.jar software. Whilst MaxEnt was originally described as a machine learning method, recent studies have shown that can be explained in terms likelihood estimation. This opens using with new settings and options, such model selection assessment criteria, improved user control variable process. New tools are needed to explore opportunities assess if they enhance performance ecological interpretability models. We present conceptual framework, Modular functionally Integrated component-based Approach (MIA) framework modelling which core components DM process decoupled then wrapped together more flexibly into functional modules. Computational object-oriented workflow approaches integrated ecological, statistical theory order handle complexity associated full way. Objects (variables, functions, results, etc.) defined according specific parameters. Properties (e.g., identities content) inherited between objects created flexible automated, yet traceable operationalise this MIA Toolbox (MIAT), set flexible, modular R-scripts (available supplementary appendices) around existing R-functions. MIAT covers range options implementation provide guidance users through A trail models increasing is built traceability interpretability, suit different purposes. briefly outline research questions addressed MIAT.