Energy distance and kernel mean embeddings for two-sample survival testing

Abstract

We study the comparison problem of distribution equality between two random samples under a right censoring scheme. To address this problem, we design a series of tests based on energy distance and kernel mean embeddings. We calibrate our tests using permutation methods and prove that they are consistent against all fixed continuous alternatives. To evaluate our proposed tests, we simulate survival curves from previous clinical trials. Additionally, we provide practitioners with a set of recommendations on how to select parameters/distances for the delay effect problem. Based on the method for parameter tunning that we propose, we show that our tests demonstrate a considerable gain of statistical power against classical survival tests.

References in corpus (6)

• Equivalence of distance-based and RKHS-based statistics in hypothesis testing
• DISCO analysis: A nonparametric extension of analysis of variance
• Nonparametric incidence estimation and bootstrap bandwidth selection in mixture cure models
• Large sample analysis of the median heuristic
• Bayesian Learning of Kernel Embeddings
• Kernel based method for the $k$-sample problem