Study of the functional role of synaptic plasticity in artificial neural networks
Lead Research Organisation:
Imperial College London
Department Name: Computing
Abstract
Study of the functional role of synaptic plasticity in artificial neural networks.
My PhD research is concerned with tackling the catastrophic forgetting problem in neural networks. In particular, I am taking inspiration from synaptic mechanisms in the brain and adapting them to improve memory in artificial neural networks that are trained on datasets that change over time
My PhD research is concerned with tackling the catastrophic forgetting problem in neural networks. In particular, I am taking inspiration from synaptic mechanisms in the brain and adapting them to improve memory in artificial neural networks that are trained on datasets that change over time
Organisations
People |
ORCID iD |
| Murray Shanahan (Primary Supervisor) | |
| Chirstos Kaplanis (Student) |
Studentship Projects
| Project Reference | Relationship | Related To | Start | End | Student Name |
|---|---|---|---|---|---|
| EP/N509486/1 | 01/10/2016 | 31/03/2022 | |||
| 1791068 | Studentship | EP/N509486/1 | 01/10/2016 | 31/03/2020 | Chirstos Kaplanis |
| Description | The goal of my research was to contribute to the field of continual learning in neural networks. Neural networks are a machine learning method that have been responsible for many recent successes in artificial intelligence but they have a number of deficiencies - one of them is that they have a poor memory when exposed to environments that change over time, which is something that must be addressed if we want neural networks to be applied in real world applications. In my research I developed three methods for addressing this problem in situations where the changes to the environment are unknown to the algorithm in advance. |
| Exploitation Route | To further inspire better algorithms for continual learning in neural networks. |
| Sectors | Other |