Fans of risk classification procedures will find this draft containing a basic – though A-to-Z complete – tutorial on how to use R for credit management useful at worst. This preliminary version of a fully fledged publication (article or book is still to be determined) contains all the relevant R code to produce enhanced visualizations as well as meaningful statistics like test error rates after cross validation, dendrograms of rating classes clustering, etc.
Though the text requires thorough elaboration to make it self explanatory, the code – edited and compiled with the package knitr on RStudio to ensure it is fully reproducible – is a comprehensive essay of credit risk classification procedures in terms of breadth and depth of discussion ranging from data preparation (it supports hand-outs for training participants as loan officer and risk managers in some major Italian banks) to findings interpretation.
Leave a Reply