A general framework for inference on algorithm-agnostic variable importance
This addresses the need for more reliable and interpretable feature importance assessment in predictive modeling, particularly in fields like healthcare, though it is incremental as it builds on existing variable importance concepts.
The authors tackled the problem of assessing variable importance beyond specific prediction algorithms by proposing a general framework for nonparametric inference on algorithm-agnostic variable importance, which allows for valid confidence intervals and hypothesis testing, as demonstrated through simulations with good operating characteristics and an application to HIV-1 antibody study data.
In many applications, it is of interest to assess the relative contribution of features (or subsets of features) toward the goal of predicting a response -- in other words, to gauge the variable importance of features. Most recent work on variable importance assessment has focused on describing the importance of features within the confines of a given prediction algorithm. However, such assessment does not necessarily characterize the prediction potential of features, and may provide a misleading reflection of the intrinsic value of these features. To address this limitation, we propose a general framework for nonparametric inference on interpretable algorithm-agnostic variable importance. We define variable importance as a population-level contrast between the oracle predictiveness of all available features versus all features except those under consideration. We propose a nonparametric efficient estimation procedure that allows the construction of valid confidence intervals, even when machine learning techniques are used. We also outline a valid strategy for testing the null importance hypothesis. Through simulations, we show that our proposal has good operating characteristics, and we illustrate its use with data from a study of an antibody against HIV-1 infection.