Bayesian techniques for Learning Invariances

Lead Research Organisation: Imperial College London
Department Name: Computing

Abstract

Invariances are a strong and ubiquitous technique to improve generalisation in machine learning methods. They are currently hand-designed, and in this project we aim to use Bayesian model selection to automatically learn them with less human input.
Research area: Medical imaging

Publications

10 25 50

Studentship Projects

Project Reference Relationship Related To Start End Student Name
EP/T51780X/1 01/10/2020 30/09/2025
2486621 Studentship EP/T51780X/1 01/10/2020 31/03/2024 Artem Artemev