A maximum-mean-discrepancy goodness-of-fit test for censored data

We introduce a kernel-based goodness-of-fit test for censored data, where observations may be missing in random time intervals: a common occurrence in clinical trials and industrial life-testing. The test statistic is straightforward to compute, as is the test threshold, and we establish consistency under the null. Unlike earlier approaches such as the Log-rank test, we make no assumptions as to how the data distribution might differ from the null, and our test has power against a very rich class of alternatives. In experiments, our test outperforms competing approaches for periodic and Weibull hazard functions (where risks are time dependent), and does not show the failure modes of tests that rely on user-defined features. Moreover, in cases where classical tests are provably most powerful, our test performs almost as well, while being more general.

[1]  Alexander J. Smola,et al.  Generative Models and Model Criticism via Optimized Maximum Mean Discrepancy , 2016, ICLR.

[2]  Sarah Friedrich,et al.  More powerful logrank permutation tests for two-sample survival data , 2018, Journal of Statistical Computation and Simulation.

[3]  Bernhard Schölkopf,et al.  Hilbert Space Embeddings and Metrics on Probability Measures , 2009, J. Mach. Learn. Res..

[4]  Ing Rj Ser.  Approximation Theorems of Mathematical Statistics , 1980 .

[5]  Aglaia Kalamatianou,et al.  A goodness of fit test for the survival function under random right censoring , 2013 .

[6]  Jun Yan.  Survival Analysis: Techniques for Censored and Truncated Data , 2004 .

[7]  E. Kaplan,et al.  Nonparametric Estimation from Incomplete Observations , 1958 .

[8]  N. Breslow,et al.  Analysis of Survival Data under the Proportional Hazards Model , 1975 .

[9]  Michael G. Akritas,et al.  The central limit theorem under censoring , 2000 .

[10]  D. Harrington.  A class of rank test procedures for censored survival data , 1982 .

[11]  Ulrich Pötter.  Covariate Effects in Periodic Hazard Rate Models , 2011 .

[12]  Kenji Fukumizu,et al.  Equivalence of distance-based and RKHS-based statistics in hypothesis testing , 2012, ArXiv.

[13]  A. Berlinet,et al.  Reproducing kernel Hilbert spaces in probability and statistics , 2004 .

[14]  David S. Moore,et al.  Chi-Square Tests of Fit for Type II Censored Data , 1980 .

[15]  A. Janssen,et al.  Weighted Logrank Permutation Tests for Randomly Right Censored Life Science Data , 2014 .

[16]  Bernhard Schölkopf,et al.  A Kernel Two-Sample Test , 2012, J. Mach. Learn. Res..

[17]  Winfried Stute,et al.  The Central Limit Theorem Under Random Censorship , 1995 .

[18]  Myles Hollander,et al.  A Chi-Squared Goodness-of-Fit Test for Randomly Censored Data , 1992 .

[19]  Kenji Fukumizu,et al.  A Linear-Time Kernel Goodness-of-Fit Test , 2017, NIPS.

[20]  Arthur Gretton,et al.  A Kernel Test of Goodness of Fit , 2016, ICML.

[21]  Krishnakumar Balasubramanian,et al.  On the Optimality of Kernel-Embedding Based Goodness-of-Fit Tests , 2017, J. Mach. Learn. Res..

[22]  N. Henze,et al.  A consistent test for multivariate normality based on the empirical characteristic function , 1988 .

[23]  Maria L. Rizzo,et al.  A new test for multivariate normality , 2005 .

[24]  Le Song,et al.  A Hilbert Space Embedding for Distributions , 2007, Discovery Science.

[25]  Mai Zhou,et al.  Combined multiple testing by censored empirical likelihood , 2009 .

[26]  Qiang Liu,et al.  A Kernelized Stein Discrepancy for Goodness-of-fit Tests , 2016, ICML.

[27]  M. Akritas.  Pearson-Type Goodness-of-Fit Tests: The Univariate Case , 1988 .

[28]  Zaïd Harchaoui,et al.  Testing for Homogeneity with Kernel Fisher Discriminant Analysis , 2007, NIPS.

[29]  M. Hollander,et al.  Testing to Determine the Underlying Distribution Using Randomly Censored Data. , 1979 .

[30]  H. Dehling,et al.  Random quadratic forms and the bootstrap for U -statistics , 1994 .

[31]  Arthur Gretton,et al.  Interpretable Distribution Features with Maximum Testing Power , 2016, NIPS.