AI Bibliography |
![]() |
Allen, C., Balavzevi'c, I., & Hospedales, T. (2020). A probabilistic framework for discriminative and neuro-symbolic semi-supervised learning. arXiv preprint arXiv:2006.05896. |
Resource type: Journal Article BibTeX citation key: Allen2020 View all bibliographic details |
Categories: Artificial Intelligence, Biological Science, Cognitive Science, Complexity Science, Computer Science, Data Sciences, Decision Theory, General, Mathematics, Neuroscience Subcategories: Autonomous systems, Decision making, Deep learning, Edge AI, Machine intelligence, Machine learning, Markov models, Neurosymbolic, Synthetic intelligence Creators: Allen, Balavzevi'c, Hospedales Publisher: Collection: arXiv preprint arXiv:2006.05896 |
Attachments |
Abstract |
In semi-supervised learning (SSL), a rule to predict labels y for data x is learned from labelled data (xl,yl) and unlabelled samples xu. Strong progress has been made by combining a variety of methods, some of which pertain to p(x), e.g. data augmentation that generates artificial samples from true x; whilst others relate to model outputs p(y|x), e.g. regularising predictions on unlabelled data to minimise entropy or induce mutual exclusivity. Focusing on the latter, we fill a gap in the standard text by introducing a unifying probabilistic model for discriminative semi-supervised learning, mirroring that for classical generative methods. We show that several SSL methods can be theoretically justified under our model as inducing approximate priors over predicted parameters of p(y|x). For tasks where labels represent binary attributes, our model leads to a principled approach to neuro-symbolic SSL, bridging the divide between statistical learning and logical rules.
|