Error-based Knockoffs Inference for Controlled Feature Selection
This work addresses controlled feature selection for researchers and practitioners in statistics and machine learning, offering a more flexible alternative to model-X knockoffs with theoretical guarantees on error rates.
The paper tackles controlled feature selection in high-dimensional finite-sample settings by proposing an error-based knockoff inference method that integrates knockoff features, error-based feature importance statistics, and a stepdown procedure, achieving competitive performance on simulated and real data.
Recently, the scheme of model-X knockoffs was proposed as a promising solution to address controlled feature selection under high-dimensional finite-sample settings. However, the procedure of model-X knockoffs depends heavily on the coefficient-based feature importance and only concerns the control of false discovery rate (FDR). To further improve its adaptivity and flexibility, in this paper, we propose an error-based knockoff inference method by integrating the knockoff features, the error-based feature importance statistics, and the stepdown procedure together. The proposed inference procedure does not require specifying a regression model and can handle feature selection with theoretical guarantees on controlling false discovery proportion (FDP), FDR, or k-familywise error rate (k-FWER). Empirical evaluations demonstrate the competitive performance of our approach on both simulated and real data.