Favourable Missingness in Semi-Supervised Classification for Exponential Mixture Models
This work studies a different regime in which the probability of label missingness depends on posterior classification uncertainty, so that the observed missing-label indicators can themselves carry information about the Bayes decision boundary.