3 papers
cs.LG2025
Monte Carlo Functional Regularisation for Continual Learning
Pengcheng Hao, Menghao Waiyan William Zhu, Ercan Engin Kuruoglu
Continual learning (CL) is crucial for the adaptation of neural network models to new environments. Although outperforming weight-space regularisation approaches, the functional re…
cs.LG2025
Sequential Function-Space Variational Inference via Gaussian Mixture Approximation
Menghao Waiyan William Zhu, Pengcheng Hao, Ercan Engin Kuruoğlu
Continual learning in neural networks aims to learn new tasks without forgetting old tasks. Sequential function-space variational inference (SFSVI) uses a Gaussian variational dist…
cs.LG2024
On Sequential Maximum a Posteriori Inference for Continual Learning
Menghao Waiyan William Zhu, Ercan Engin Kuruoğlu
We formulate sequential maximum a posteriori inference as a recursion of loss functions and reduce the problem of continual learning to approximating the previous loss function. We…