Mingtian Zhang
PhD student in machine learning
Centre for Artificial Intellige, University College London
I am a PhD student in UCL supervised by Prof. David Barber and Prof. Brooks Paige.
I am interested in methodologies and applications of probabilistic modelling.
Blog Posts
- A failure case of VAE Dec 31, 2021
Selected Publications and Preprints
-
Spread Flows for Manifold Modelling
AISTATS 2023
Mingtian Zhang, Yitong Sun, Steven McDonagh and Chen Zhang -
Generalization Gap in Amortized Inference
NeurIPS 2022
Mingtian Zhang, Peter Hayes and David Barber -
Towards Healing the Blindness of Score Matching
Score-Based Methods Workshop, NeurIPS 2022
Mingtian Zhang, Oscar Keys, Peter Hayes, David Barber, Brooks Paige and François-Xavier Briol -
Out-of-Distribution Detection with Class Ratio Estimation
Machine Learning Safety Workshop, NeurIPS 2022
Mingtian Zhang, Andi Zhang, Tim Z. Xiao, Yitong Sun and Steven McDonagh -
Integrated Weak Learning
ArXiv preprint
Peter Hayes, Mingtian Zhang, Raza Habib, Jordan Burgess, Emine Yilmaz and David Barber -
Improving VAE-based Representation Learning
ArXiv preprint
Mingtian Zhang, Tim Xiao, Brooks Paige and David Barber -
Parallel Neural Local Lossless Compression
ArXiv preprint
Mingtian Zhang, James Townsend, Ning Kang and David Barber -
On the Out-of-Distribution Generalization
of Probabilistic Image Modelling
NeurIPS 2021
Mingtian Zhang, Andi Zhang and Steven McDonagh -
Active Forgetting of Negative Transfer in Continual Learning
NeurIPS 2021
Liyuan Wang, Mingtian Zhang, Zhongfan Jia, Qian Li, Kaisheng Ma, Chenglong Bao, Jun Zhu and Yi Zhong -
Solipsistic Reinforcement Learning
Self-Supervision for Reinforcement Learning Workshop, ICLR 2021
Mingtian Zhang, Peter Hayes, Tim Xiao, Andi Zhang and David Barber -
Spread Divergence
ICML 2020
Mingtian Zhang, Peter Hayes, Thomas Bird, Raza Habib and David Barber -
Variational f-Divergence Minimization
Information Theory and Machine Learning Workshop, NeurIPS 2019
Mingtian Zhang, Thomas Bird, Raza Habib, Tianlin Xu and David Barber
Softwares
- NeLLoC: A CPU-based light-weight neural lossless image compressor that can compress image with arbitrary sizes.
- ParallelNeLLoC: A parallel implementation of NeLLoC with big and free speed improvement.