Are We Ready for Semi-Supervised Learning for Big Data? From Theory to Practice

OVERVIEW

The last years, machine learning has experienced an astonishing development in all domains. It has become the technique to-go for most of existing problems reporting state-of-the-art results in tasks such as classification, segmentation and detection. However, a major drawback of these techniques is the high dependence on a large corpus of labelled data. This might be a strong assumption for a solution, as annotated data contains strong human bias, and for many applications it is expensive and time consuming to obtain labels. Motivated by these drawbacks, different philosophies have experience a fast developments including Semi-Supervised Learning, in which one aims to exploit very small set of label data and a huge amount of unlabelled data. In this session, we aim to discuss the recent developments on semi-supervised learning for real-world problems. In particular, we will highlight the current state-of-the-art techniques, challenges and opportunities. We will address these factors from both theoretical and practical points of view.