arXiv subjects

Wanchuang Zhu

Publications and source records attributed to Wanchuang Zhu.

6 recordsLinked to original sources

arXiv2026

MANDERA: Malicious Node Detection in Federated Learning via Ranking

Byzantine attacks hinder the deployment of federated learning algorithms. Although we know that the benign gradients and Byzantine attacked gradients are distributed differently, to detect the malicious gradients is challenging due to (1) the gradient is high-dimensional and each dimension has its unique distribution and (2) the benign gradients and the attacked gradients are always mixed (two-sample test methods cannot apply directly). To address the above, for the first time, we propose MANDERA which is theoretically guaranteed to efficiently detect all malicious gradients under Byzantine attacks with no prior knowledge or history about the number of attacked nodes. More specifically, we transfer the original updating gradient space into a ranking matrix. By such an operation, the scales of different dimensions of the gradients in the ranking space become identical. The high-dimensional benign gradients and the malicious gradients can be easily separated. The effectiveness of MANDERA is further confirmed by experimentation on four Byzantine attack implementations (Gaussian, Zero Gradient, Sign Flipping, Shifted Mean), comparing with state-of-the-art defenses. The experiments cover both IID and Non-IID datasets.

cs.LG↗

arXiv2022

Structure Learning for Hybrid Bayesian Networks

Bayesian networks have been used as a mechanism to represent the joint distribution of multiple random variables in a flexible yet interpretable manner. One major challenge in learning the structure of a Bayesian network is how to model networks which include a mixture of continuous and discrete random variables, known as hybrid Bayesian networks. This paper overviews the literature on approaches to handle hybrid Bayesian networks. Typically one of two approaches is taken: either the data are considered to have a joint distribution which is designed for a mixture of discrete and continuous variables, or continuous random variables are discretized, resulting in discrete Bayesian networks. In this paper, we propose a strategy to model all random variables as Gaussian, referred to it as {\it Run it As Gaussian (RAG)}. We demonstrate that RAG results in more reliable estimates of graph structures theoretically and by simulation studies, than converting continuous random variables to discrete. Both strategies are also implemented on a childhood obesity data set. The two different strategies give rise to significant differences in the optimal graph structures, with the results of the simulation study suggesting that our strategy is more reliable.

stat.ME↗

arXiv2021

Partition-Mallows Model and Its Inference for Rank Aggregation

Learning how to aggregate ranking lists has been an active research area for many years and its advances have played a vital role in many applications ranging from bioinformatics to internet commerce. The problem of discerning reliability of rankers based only on the rank data is of great interest to many practitioners, but has received less attention from researchers. By dividing the ranked entities into two disjoint groups, i.e., relevant and irrelevant/background ones, and incorporating the Mallows model for the relative ranking of relevant entities, we propose a framework for rank aggregation that can not only distinguish quality differences among the rankers but also provide the detailed ranking information for relevant entities. Theoretical properties of the proposed approach are established, and its advantages over existing approaches are demonstrated via simulation studies and real-data applications. Extensions of the proposed method to handle partial ranking lists and conduct covariate-assisted rank aggregation are also discussed.

stat.ME↗

arXiv2016

A Bayesian spatial temporal mixtures approach to kinetic parametric images in dynamic Positron Emission Tomography

We present a fully Bayesian statistical approach to the problem of compartmental modelling in the context of Positron Emission Tomography. We cluster homogeneous region of interest and perform kinetic parameter estimation simultaneously. A mixture modelling approach is adopted, incorporating both spatial and temporal information based on reconstructed dynamic PET image. Our modelling approach is flexible, and provides uncertainty estimates for the estimated kinetic parameters. Crucially, the proposed method allows us to determine the unknown number of clusters, which has a great impact on resulting estimated kinetic parameters. We demonstrate our method on simulated dynamic Myocardial PET data, and show that our method is superior to standard curve-fitting approach.

stat.AP↗

arXiv2016

A novel approach for Markov Random Field with intractable normalising constant on large lattices

The pseudo likelihood method of Besag(1974), has remained a popular method for estimating Markov random field on a very large lattice, despite various documented deficiencies. This is partly because it remains the only computationally tractable method for large lattices. We introduce a novel method to estimate Markov random fields defined on a regular lattice. The method takes advantage of conditional independence structures and recursively decomposes a large lattice into smaller sublattices. An approximation is made at each decomposition. Doing so completely avoids the need to compute the troublesome normalising constant. The computational complexity is $O(N)$, where $N$ is the the number of pixels in lattice, making it computationally attractive for very large lattices. We show through simulation, that the proposed method performs well, even when compared to the methods using exact likelihoods.

stat.ME↗

arXiv2014

Relabelling Algorithms for Large Dataset Mixture Models

Mixture models are flexible tools in density estimation and classification problems. Bayesian estimation of such models typically relies on sampling from the posterior distribution using Markov chain Monte Carlo. Label switching arises because the posterior is invariant to permutations of the component parameters. Methods for dealing with label switching have been studied fairly extensively in the literature, with the most popular approaches being those based on loss functions. However, many of these algorithms turn out to be too slow in practice, and can be infeasible as the size and dimension of the data grow. In this article, we review earlier solutions which can scale up well for large data sets, and compare their performances on simulated and real datasets. In addition, we propose a new, and computationally efficient algorithm based on a loss function interpretation, and show that it can scale up well in larger problems. We conclude with some discussions and recommendations of all the methods studied.

stat.AP↗