Reliable Semi-Supervised Learning when Labels are Missing at Random
This addresses the issue of performance degradation in semi-supervised learning for practitioners needing reliable uncertainty estimates when labels are missing at random.
The paper tackles the problem of unreliable classifiers in semi-supervised learning due to restrictive assumptions about unlabeled features, developing an approach that provides reliable uncertainty quantification for labels missing at random, as demonstrated on handwritten digit and cloth classification datasets.
Semi-supervised learning methods are motivated by the availability of large datasets with unlabeled features in addition to labeled data. Unlabeled data is, however, not guaranteed to improve classification performance and has in fact been reported to impair the performance in certain cases. A fundamental source of error arises from restrictive assumptions about the unlabeled features, which result in unreliable classifiers that underestimate their prediction error probabilities. In this paper, we develop a semi-supervised learning approach that relaxes such assumptions and is capable of providing classifiers that reliably quantify the label uncertainty. The approach is applicable using any generative model with a supervised learning algorithm. We illustrate the approach using both handwritten digit and cloth classification data where the labels are missing at random.