Rui-Yang Zhang
@ryzhang.bsky.social
PhD student in Computational Statistics and Machine Learning at STOR-i CDT, Lancaster University, UK.
Research Interests: Sampling Algorithms, Bayesian Experiment Designs, Neural Amortization.
https://shusheng3927.github.io/
Research Interests: Sampling Algorithms, Bayesian Experiment Designs, Neural Amortization.
https://shusheng3927.github.io/
Preferential Sampling refers to scenarios where observation locations are confounded by the field of interest which the same observations are used to infer. This recent arxiv (arxiv.org/abs/2511.03158) looked at how harmful ignoring preferential sampling would be - not much, according to the paper.
November 7, 2025 at 12:52 PM
Preferential Sampling refers to scenarios where observation locations are confounded by the field of interest which the same observations are used to infer. This recent arxiv (arxiv.org/abs/2511.03158) looked at how harmful ignoring preferential sampling would be - not much, according to the paper.
Is it just me or does Google Scholar forbid searches via Avanti’s WiFi?
June 9, 2025 at 9:05 AM
Is it just me or does Google Scholar forbid searches via Avanti’s WiFi?
Yes! Their fig1b shown below is rather surprising: when well-tuned, IMH (with NF proposal) ≈ Jump MH ≈ HMC ≈ NeuTra HMC.
I guess this also indicates how truly great HMC is: even with all those NN-aided NF help, it still can't quite beat the good-old Hamiltonian ...
I guess this also indicates how truly great HMC is: even with all those NN-aided NF help, it still can't quite beat the good-old Hamiltonian ...
December 26, 2024 at 1:44 PM
Yes! Their fig1b shown below is rather surprising: when well-tuned, IMH (with NF proposal) ≈ Jump MH ≈ HMC ≈ NeuTra HMC.
I guess this also indicates how truly great HMC is: even with all those NN-aided NF help, it still can't quite beat the good-old Hamiltonian ...
I guess this also indicates how truly great HMC is: even with all those NN-aided NF help, it still can't quite beat the good-old Hamiltonian ...
Inspiring paper on empirical evaluations and benchmarking of Normalising Flow-based MCMC algorithms with standard MCMC!
arxiv: arxiv.org/abs/2412.17136
arxiv: arxiv.org/abs/2412.17136
December 26, 2024 at 1:13 PM
Inspiring paper on empirical evaluations and benchmarking of Normalising Flow-based MCMC algorithms with standard MCMC!
arxiv: arxiv.org/abs/2412.17136
arxiv: arxiv.org/abs/2412.17136
