Variational inference based unsupervised continual learning
This research is aimed at investigating variational inference based deep learning approach for generative continual learning. Continual learning is aimed at learning a sequence of task, in scenarios where data from past tasks are unavailable. Thus, it emphasizes on learning a sequence of task, witho...
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Thesis-Master by Coursework |
| Language: | English |
| Published: |
Nanyang Technological University
2022
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/155840 |
| _version_ | 1811691599678144512 |
|---|---|
| author | Gao, Zhaoqi |
| author2 | Ponnuthurai Nagaratnam Suganthan |
| author_facet | Ponnuthurai Nagaratnam Suganthan Gao, Zhaoqi |
| author_sort | Gao, Zhaoqi |
| collection | NTU |
| description | This research is aimed at investigating variational inference based deep learning approach for generative continual learning. Continual learning is aimed at learning a sequence of task, in scenarios where data from past tasks are unavailable. Thus, it emphasizes on learning a sequence of task, without catastrophically forgetting any task. Bayesian deep neural networks enable inherent online variational inference and are suitable candidates for continual learning. Studies in the literature have shown the continual learning ability of Bayesian neural network in a sequence of classification task.
In this thesis, we develop an unsupervised variational inference based continual learning algorithm with generative replay, using variational autoencoders. We demonstrate the performance of the proposed approach on split MNIST and split CIFAR10 data sets. Performances studies in comparison with state-of-the-art continual learning approaches show improved performance of the proposed approach with unsupervised generative replay. We also define some criteria to evaluate the performance of an unsupervised continual learning model. |
| first_indexed | 2024-10-01T06:22:27Z |
| format | Thesis-Master by Coursework |
| id | ntu-10356/155840 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2024-10-01T06:22:27Z |
| publishDate | 2022 |
| publisher | Nanyang Technological University |
| record_format | dspace |
| spelling | ntu-10356/1558402023-07-04T17:43:25Z Variational inference based unsupervised continual learning Gao, Zhaoqi Ponnuthurai Nagaratnam Suganthan School of Electrical and Electronic Engineering A*STAR EPNSugan@ntu.edu.sg Engineering::Electrical and electronic engineering This research is aimed at investigating variational inference based deep learning approach for generative continual learning. Continual learning is aimed at learning a sequence of task, in scenarios where data from past tasks are unavailable. Thus, it emphasizes on learning a sequence of task, without catastrophically forgetting any task. Bayesian deep neural networks enable inherent online variational inference and are suitable candidates for continual learning. Studies in the literature have shown the continual learning ability of Bayesian neural network in a sequence of classification task. In this thesis, we develop an unsupervised variational inference based continual learning algorithm with generative replay, using variational autoencoders. We demonstrate the performance of the proposed approach on split MNIST and split CIFAR10 data sets. Performances studies in comparison with state-of-the-art continual learning approaches show improved performance of the proposed approach with unsupervised generative replay. We also define some criteria to evaluate the performance of an unsupervised continual learning model. Master of Science (Computer Control and Automation) 2022-03-23T23:41:08Z 2022-03-23T23:41:08Z 2021 Thesis-Master by Coursework Gao, Z. (2021). Variational inference based unsupervised continual learning. Master's thesis, Nanyang Technological University, Singapore. https://hdl.handle.net/10356/155840 https://hdl.handle.net/10356/155840 en application/pdf Nanyang Technological University |
| spellingShingle | Engineering::Electrical and electronic engineering Gao, Zhaoqi Variational inference based unsupervised continual learning |
| title | Variational inference based unsupervised continual learning |
| title_full | Variational inference based unsupervised continual learning |
| title_fullStr | Variational inference based unsupervised continual learning |
| title_full_unstemmed | Variational inference based unsupervised continual learning |
| title_short | Variational inference based unsupervised continual learning |
| title_sort | variational inference based unsupervised continual learning |
| topic | Engineering::Electrical and electronic engineering |
| url | https://hdl.handle.net/10356/155840 |
| work_keys_str_mv | AT gaozhaoqi variationalinferencebasedunsupervisedcontinuallearning |