References & Citations
Statistics > Methodology
Title: Interpreting Models via Single Tree Approximation
(Submitted on 27 Oct 2016)
Abstract: We propose a procedure to build a decision tree which approximates the performance of complex machine learning models. This single approximation tree can be used to interpret and simplify the predicting pattern of random forests (RFs) and other models. The use of a tree structure is particularly relevant in medical questionnaires where it enables an adaptive shortening of the questionnaire, reducing response burden. We study the asymptotic behavior of splits and introduce an improved splitting method designed to stabilize tree structure. Empirical studies on both simulation and real data sets illustrate that our method can simultaneously achieve high approximation power and stability.
Link back to: arXiv, form interface, contact.