Related projects
Discover more projects across a range of sectors and discipline — from AI to cleantech to social innovation.
Deep neural network (DNN) is a class of machine learning algorithms which is inspired by biological neural networks. DNNs are themselves general function approximations, which is the reason they can be applied to almost any machine learning problem. Their applications can be found in visual object recognition in computer vision, translating texts in unsupervised learning, etc. DNNs are prone to overfitting because DNNs usually have many more parameters than the available training data. However, they usually have a low error on the test data. This surprising fact has motivated the scientific community to study the generalization performance of DNNs. Nevertheless, the previous attempts do not lead to a satisfying answer to the aforementioned question. In this project, we aim to introduce an information-theoretic framework which let us find a promising answer to the question of why DNNs generalize well in practice.
Daniel Roy
Mahdi Haghifam
Element AI
Visual arts
Information and communications technologies
Accelerate
Discover more projects across a range of sectors and discipline — from AI to cleantech to social innovation.
Find the perfect opportunity to put your academic skills and knowledge into practice!
Find ProjectsThe strong support from governments across Canada, international partners, universities, colleges, companies, and community organizations has enabled Mitacs to focus on the core idea that talent and partnerships power innovation — and innovation creates a better future.