… standard multiple testing procedures is that the null distribution should be known. Here,
we consider a null distribution-free approach for multiple testing in the following semisupervised …

… Based on the previous ideas and feature extraction (FE) schemes, we investigate the
possibility of using a semi-supervised model selection algorithm based on hypothesis testing …

… connections to multiple testing, where we show that SSND … multiple testing literature. The
SSND model, and the results presented in this paper, are thus relevant to multiple testing as …

… Chapter 10 reviews the results of this thesis, and provides a guide for practitioners on
hypothesis testing, effect size estimation and feature selection in semisupervised and positive-…

… the hypotheses according to how confident we believe each hypothesis is a false null. Indeed,
… In this paper, we offer a novel approach for multiple testing with side-information. The idea …

… boosting algorithms have been developed based on semi-supervised smoothness and …
[12] to a semisupervised boosting framework with regularization working on semi-supervised …

… Firstly we will review information theoretic feature selection, via hypothesis testing and
ranking. Then we will formally introduce the semi-supervised settings that we will focus on, and …

… Semi-supervised learning has proven to be a powerful paradigm for leveraging unlabeled
… In this work, we unify the current dominant approaches for semi-supervised learning to …

… of software testing is deciding whether a test case has … semi-supervised learning on dynamic
execution data (test inputs/outputs and execution traces). A small proportion of the test data …

… to define the task of semi-supervised multi-target regression. We formalize it as follows. … We
denote semi-supervised PCTs and semi-supervised random forests where feature weighting …