Interactive graphics for visually diagnosing forest classifiers in R

This article describes structuring data and constructing plots to explore forest classification models interactively. A forest classifier is an example of an ensemble since it is produced by bagging multiple trees. The process of bagging and combining results from multiple trees produces numerous diagnostics which, with interactive graphics, can provide a lot of insight into class structure in high dimensions. Various aspects of models are explored in this article, to assess model complexity, individual model contributions, variable importance and dimension reduction, and uncertainty in prediction associated with individual observations. The ideas are applied to the random forest algorithm and projection pursuit forest but could be more broadly applied to other bagged ensembles helping in the interpretability deficit of these methods. Interactive graphics are built in R using the ggplot2, plotly, and shiny packages.

Automated summary

This article explores the use of interactive graphics to analyze forest classification models. By combining and analyzing the results from multiple decision trees, a forest classifier can provide insights into class structure in high dimensions. The article discusses various aspects of the models, including model complexity, variable importance, and prediction uncertainty. The methods described can be applied to random forest and other bagged ensemble models, helping to improve their interpretability. The graphics are built using the ggplot2, plotly, and shiny packages in R.