More Than Meets The Eye: Semi-supervised Learning Under Non-IID Data
This addresses a practical issue for researchers and practitioners in semi-supervised learning, but it is incremental as it builds on existing methods with a new matching approach.
The paper tackles the problem of selecting unlabeled data for semi-supervised deep learning under non-IID conditions, showing that semantic matching can degrade performance and proposing density-based dissimilarity measures as a more reliable criterion.
A common heuristic in semi-supervised deep learning (SSDL) is to select unlabelled data based on a notion of semantic similarity to the labelled data. For example, labelled images of numbers should be paired with unlabelled images of numbers instead of, say, unlabelled images of cars. We refer to this practice as semantic data set matching. In this work, we demonstrate the limits of semantic data set matching. We show that it can sometimes even degrade the performance for a state of the art SSDL algorithm. We present and make available a comprehensive simulation sandbox, called non-IID-SSDL, for stress testing an SSDL algorithm under different degrees of distribution mismatch between the labelled and unlabelled data sets. In addition, we demonstrate that simple density based dissimilarity measures in the feature space of a generic classifier offer a promising and more reliable quantitative matching criterion to select unlabelled data before SSDL training.