Supervised neural networks have been applied as a machine learning technique to identify and predict emergent patterns among multiple variables. A common criticism of these methods is the inability to characterize relationships among variables from a fitted model. Although several techniques have been proposed to "illuminate the black box", they have not been made available in an open-source programming environment. This article describes the NeuralNetTools package that can be used for the interpretation of supervised neural network models created in R. Functions in the package can be used to visualize a model using a neural network interpretation diagram, evaluate variable importance by disaggregating the model weights, and perform a sensitivity analysis of the response variables to changes in the input variables. Methods are provided for objects from many of the common neural network packages in R, including caret, neuralnet, nnet, and RSNNS. The article provides a brief overview of the theoretical foundation of neural networks, a description of the package structure and functions, and an applied example to provide a context for model development with NeuralNetTools. Overall, the package provides a toolset for neural networks that complements existing quantitative techniques for data-intensive exploration.
Artificial neural networks (ANNs) are powerful tools for data analysis and are particularly suitable for modeling relationships between variables for best prediction of an outcome. While these models can be used to answer many important research questions, their utility has been critically limited because the interpretation of the "black box" model is difficult. Clinical investigators usually employ ANN models to predict the clinical outcomes or to make a diagnosis; the model however is difficult to interpret for clinicians. To address this important shortcoming of neural network modeling methods, we describe several methods to help subject-matter audiences (e.g., clinicians, medical policy makers) understand neural network models. Garson's algorithm describes the relative magnitude of the importance of a descriptor (predictor) in its connection with outcome variables by dissecting the model weights. The Lek's profile method explores the relationship of the outcome variable and a predictor of interest, while holding other predictors at constant values (e.g., minimum, 20th quartile, maximum). While Lek's profile was developed specifically for neural networks, partial dependence plot is a more generic version that visualize the relationship between an outcome and one or two predictors. Finally, the local interpretable model-agnostic explanations (LIME) method can show the predictions of any classification or regression, by approximating it locally with an interpretable model. R code for the implementations of these methods is shown by using example data fitted with a standard, feed-forward neural network model. We offer codes and step-by-step description on how to use these tools to facilitate better understanding of ANN.
The economic benefits of dams have been assumed to outweigh the costs, thus providing rationale for construction of dams around the world. However, the development of these structures can be accompanied by negative biophysical, socio-economic, and geopolitical impacts; often through the loss of ecosystem services provided by fully functioning aquatic systems. Moreover, impacts of dams can be involuntarily imposed on marginalized peoples whose livelihoods are dependent on riverine resources. In this review, we examine the impacts of dam projects in regions of the world that are at different stages of development, using the USA, China, and Southeast Asia to represent a development gradient from developed to developing, respectively. Case studies for each region illustrate the environmental and livelihood impacts of dams in each region, while also providing a basis to better understand how environmental degradation is directly related to economic growth. We conclude that a distinct temporal component related to development mediates the relationship between policies and governance mechanisms and the mitigation of environmental and social costs of dams. The role of affected individuals to influence the political will behind dam projects and the importance of environmental advocacy is emphasized as a fundamental approach towards more sustainable development.
Biological invasions are projected to be the main driver of biodiversity and ecosystem function loss in lakes in the 21st century. However, the extent of these future losses is difficult to quantify because most invasions are recent and confounded by other stressors. In this study, we quantified the outcome of a century-old invasion, the introduction of common carp to North America, to illustrate potential consequences of introducing non-native ecosystem engineers to lakes worldwide. We used the decline in aquatic plant richness and cover as an index of ecological impact across three ecoregions: Great Plains, Eastern Temperate Forests and Northern Forests. Using whole-lake manipulations, we demonstrated that both submersed plant cover and richness declined exponentially as carp biomass increased such that plant cover was reduced to <10% and species richness was halved in lakes in which carp biomass exceeded 190 kg ha . Using catch rates amassed from 2000+ lakes, we showed that carp exceeded this biomass level in 70.6% of Great Plains lakes and 23.3% of Eastern Temperate Forests lakes, but 0% of Northern Forests lakes. Using model selection analysis, we showed that carp was a key driver of plant species richness along with Secchi depth, lake area and human development of lake watersheds. Model parameters showed that carp reduced species richness to a similar degree across lakes of various Secchi depths and surface areas. In regions dominated by carp (e.g., Great Plains), carp had a stronger impact on plant richness than human watershed development. Overall, our analysis shows that the introduction of common carp played a key role in driving a severe reduction in plant cover and richness in a majority of Great Plains lakes and a large portion of Eastern Temperate Forests lakes in North America.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.