Title: A combined p-value test for the mean difference of high-dimensional data
Authors: Wangli Xu - Renmin University of China (China) [presenting]
Abstract: A novel method for testing the equality of high-dimensional means is proposed by employing the idea of multiple hypothesis test. The proposed statistic is the maximum of standardized partial sums of logarithmic p-values. Numerical studies show that the proposed method performs well for both normal and non-normal data. In addition, under both dense and sparse alternative hypotheses, our test can have good power performance. For illustration, a real data analysis is implemented.