Title: Consensus and Random Walks on “Higher-order Networks

Time: 10:30am 

Venue: CB308, 3/F, Chow Yei Ching Building, HKU

Speaker(s): Prof. Renaud Lambiotte

Remark(s): 

Abstract

Renaud Lambiotte is a Professor of Networks and Nonlinear Systems at the Mathematical Institute of the University of Oxford and a Tutorial Fellow at Somerville College. As a distinguished network scientist, he is also a Turing Fellow at the Alan Turing Institute in London and an External Faculty member at the Complexity Science Hub Vienna. After receiving his Ph.D. in Physics from the Université Libre de Bruxelles, he held postdoctoral positions at prestigious institutions including ENS Lyon, Université de Liège, UCLouvain, and Imperial College London before becoming a Professor of Mathematics at the University of Namur. His research primarily focuses on the modeling and analysis of large-scale networks, with particular emphasis on social networks, brain networks, network clustering algorithms, empirical hypergraphs, and temporal networks.

Professor Lambiotte has made significant contributions to the field with approximately 150 peer-reviewed publications and is the author of two books: Modularity and Dynamics on Complex Networks (Cambridge University Press, 2022) and A Guide to Temporal Networks (World Scientific,2021). He currently serves as a Senior Associate Editor for Science Advances, further demonstrating his continued influence in the academic community.

About the speaker

Random walks play a central role in modern network science, as a model for diffusion and to extract "multi-scale" information in relational data. In this talk, I will give an overview of recent generalisations to higherorder networks. These generalisations include hypergraphs, accounting for multiway interactions, temporal networks, where edges are dynamical objects, and signed networks, allowing for negative edges to encode conflictual interactions.

Title: Ambient Sensing and Intelligence: Where AI Meets Physics

Time: 03:00pm 

Venue: HW312, Haking Wong Building

Speaker(s): Prof. K. J. Ray Liu

Remark(s): 

Abstract

Utilizing ambient radio waves to monitor human activities has long been a dream of many. Historically treated as unwanted interference, the millions of bouncing radio multipaths in everyday environments can now be observed, thanks to the broader bandwidths of modern Wi-Fi, IoT, and 5G/6G devices.
This talk introduces a paradigm shift: turning such interference into intelligence by treating multipaths as hundreds, if not millions, of virtual sensors. A new physical principle will be presented, showing that the time-reversal focusing spot exhibits a stationary Bessel function power distribution, enabling accurate speed estimation even under severe non-line-of-sight conditions. Defying long-standing scientific belief, this approach thrives indoors where the Doppler Effect fails, proving that indoor environmental complexity is actually a source of precision. Enhanced by AI, this revolutionary ambient intelligence enables a new wave of device-free, non-obtrusive IoT applications. The talk will feature the world’s first centimeter-accuracy wireless indoor positioning system, alongside applications in contactless vital signs detection, sleep monitoring, and fall detection using commodity Wi-Fi. Ultimately, ambient sensing gives future wireless networks a "sixth sense" to decipher the world around us and will forever change the future of wireless systems.

About the speaker

K. J. Ray Liu is the founder of the award-winning Origin AI, acquired by ADT in 2026, that pioneers ambient sensing and intelligence. He was the 2022 IEEE President and CEO and 2012-13 President of IEEE Signal Processing Society. He retired from University of Maryland, College Park, as Distinguished University Professor. He has trained 76 doctoral/postdoctoral students, of which 14 are now IEEE fellows with over 200 doctoral descendants. Prof. Liu is a recipient of many prestigious awards, including two IEEE Technical Field Awards: the 2021 IEEE Fourier Award for Signal Processing and the 2016 IEEE Leon K. Kirchmayer Graduate Teaching Award, 2026 IEEE Haraden Pratt Award, and also IEEE Signal Processing Society 2014 Norbert Wiener Lifetime Achievement Award, 2009 Claude Shannon-Harry Nyquist Technical Achievement Award, and more than a dozen best paper awards. Recognized as a Web of Science Highly Cited Researcher, he is a member of National Academy of Engineering and a Fellow of IEEE, the American Association for the Advancement of Science (AAAS), and the National Academy of Inventors.

Title: Statistical Estimation and Inference in High-Dimensional and Semiparametric Models

Time: 10:30am 

Venue: RR301, Run Run Shaw Building

Speaker(s): Dr. Yuanhang Luo

Remark(s): 

Abstract

High-dimensional parameters arise naturally in many modern statistical problems — from regression with a large number of covariates to ranking models where the number of latent utilities grows with the size of the comparison graph. In this talk, I will present two recent works that address estimation and inference challenges in these high-dimensional and semiparametric settings.

The first part concerns online inference in high-dimensional generalized linear models with streaming data. We develop the Adaptive Debiased Lasso (ADL), which updates coefficient estimates and confidence intervals upon each new data arrival. The method features an adaptive stochastic gradient descent algorithm with a novel online debiasing procedure via Taylor approximation, achieving asymptotic normality with only O(p) space and time complexity instead of O(p²) in previous methods. In the second part, I will present a semiparametric model for ranking data whose underlying graph structure governs both the dimensionality of the problem and the difficulty of estimation. In the comparison hypergraph, each object's strength is modeled as the sum of a utility parameter and a nonparametric covariate effect approximated by a deep neural network. Non-asymptotic error bounds achieving minimax optimality for model components are established. The framework is demonstrated on an ATP tennis dataset that capturing nonlinear contextual effects in player performance.

About the speaker

Yuanhang Luo is currently a PhD student in the Department of Data Science and Artificial Intelligence at the Hong Kong Polytechnic University. He received his B.Sc. in Mathematics & Statistics from Hong Kong Baptist University. His research focuses on high-dimensional statistics, ranking and reinforcement learning.

Title: Addressing Biases, Batches and Hidden Heterogeneities in Microbiome Studies

Time: 11:00am 

Venue: Room 301, Run Run Shaw Building

Speaker(s): Prof. Ni Zhao

Remark(s): 

Abstract

Microbiome data, like other high-throughput omics data, are susceptible to technical artifacts, including batch effects, measurement biases, and latent sources of heterogeneity. These challenges present major barriers to large-scale, multi-site, and integrative microbiome studies, where existing methods often rely on restrictive assumptions and may yield unreliable inference under realistic community-level variation. In this presentation, I will highlight recent methodological advances from our group to address these challenges, including ConQuR, a method for correcting known batch effects; QuanT, a framework for detecting latent or unknown sources of heterogeneity; and CAFT, a statistically principled approach for mitigating bias in differential abundance analysis. These methods are built on flexible nonparametric statistical models that accommodate the irregular, heavy-tailed, and zero-inflated characteristics of microbiome data, enabling more robust and reliable inference across diverse study settings.

About the speaker

Dr. Ni Zhao received her PhD from the University of North Carolina at Chapel Hill and is currently an Associate Professor and PhD Program Director in the Department of Biostatistics at Johns Hopkins University. Her primary research interests lie in statistical genetics and genomics, with a particular focus on developing statistical methods for microbiome studies, including both bulk and spatial microbiome profiling. In recent years, her lab has made significant contributions to understanding and addressing batch effects, biases, and other sources of technical variation in microbiome studies. She has also been actively involved in large-scale epidemiologic studies, where microbiome and multi-omics data integration are central components. Over the past decade, Dr. Zhao has published more than 40 peer-reviewed papers in leading journals across statistics, epidemiology, and biomedical and clinical research.

Title: Evaluating Causes of Effects by Posterior Effects of Causes

Time: 10:30am 

Venue: Room 301, Run Run Shaw Building

Speaker(s): Dr. Zitong Lu

Remark(s): 

Abstract

As highlighted in Dawid (2000) and Pearl & Mackenzie (2018), deducing the causes of given effects is a more challenging problem than evaluating the effects of causes in causal inference. For the case with a single causal variable, the probability of causation and the probability of necessity have been used to assess causes of effects. For a case with multiple causes that may affect each other, we propose the posterior causal effects based on observed evidence, as a measure of causes of effects. Since posterior causal effects involve probabilities of counterfactual variables, their identifiability requires assumptions of no confounding and monotonicity beyond those needed for traditional causal effects; we present these assumptions and provide the corresponding identification equations. We further extend this framework to settings with multiple effect variables. The proposed approach applies broadly to causal attribution, medical diagnosis, and the assessment of blame and responsibility in studies with multiple effect or outcome variables, and we illustrate it through numerical examples.

About the speaker

Dr. Zitong Lu is a Postdoctoral Fellow in the Department of Statistics and Data Science at the Chinese University of Hong Kong. He received his Ph.D. in Systems Engineering from City University of Hong Kong and a B.Sc. in Statistics from Peking University. His research focuses on causal inference, particularly causal attribution and individual treatment effects.

Title: Functional Dynamics in Non-Functional Data via Principle Component Analysis

Time: 10:30am 

Venue: Room 301, Run Run Shaw Building

Speaker(s): Prof. Zhijie Xiao

Remark(s): 

Abstract

Distributions of many economic and financial time series variables contain important information that affects investment and economic policy. The study of distributional relationship has attracted a lot of research attentions recently, and many models are proposed to capture dependence on different aspects of the distributional relationship. The majority of existing literature consider this problem by focusing on some selected representative characteristics of a distribution - such as the variance, dispersion, or a particular quantile - and model dependence based on these selected characteristics. We argue that focusing only on some selected characteristics could potentially miss important information about the dependence relationship, and propose functional quantile regression models to study the distributional dependence relationship in time series data. In the proposed functional quantile regression models, the future economic behavior can be affected by the past distributional information in the economy. The models can capture systematic influences of the past distributional information on the conditional distribution of the response, and therefore constitute a significant extension of traditional time series models in which the effect of conditioning information is confined to only a few selected characteristics of the past distribution. Unlike traditional functional regression models that rely on rich data environments with functional data features, our approach focuses on functional relationships within conventional time series data. We consider a linear functional quantile autoregression model and explore estimation and dimension reduction via Functional Principal Component Analysis (FPCA) in this paper. We propose a threestep estimation procedure, and analyze limiting properties of the proposed estimators. Uniform asymptotic results are developed to facilitate statistical inference based on the functional model. We show that the proposed FPCA-based estimator of the conditional quantile function achieves near root-n convergence rate, improving upon the nonparametric rate of conventional sieve estimators. Monte Carlo experiments are conducted and show improved finite sample performance of the proposed estimator compared to other estimators. Finally, an empirical application to S&P 500 index illustrates the potential of the new method in capturing complex risk dynamics.

About the speaker

Zhijie Xiao currently is a professor at the Department of Economics, Boston College. He obtained his PhD in Economics from Yale University in 1997. His researches cover all kinds of areas in econometrics and statistics, and especially he is a leading figure in quantile regression. Prof Xiao has received many awards from econometric community, including Plura Scripsit Award in Econometric Theory, fellow of Journal of Econometrics, etc., and he has served, is serving, as the editorial board for top journals in econometrics and statistics, such as Journal of Econometric, JASA, etc.

Title: Subsystem Quantum Error Correction for Noisy Quantum Metrology

Time: 02:30pm 

Venue: HW312, Haking Wong Building, HKU

Speaker(s): Dr. Qiushi Liu

Remark(s): 

Abstract

Quantum error correction has been successfully applied to enhance the precision of parameter estimation in the presence of noise. Nonetheless, existing methods require a number of noiseless, controllable ancillae and lack efficient encoding and decoding procedures. In this Letter, we demonstrate that subsystem error correction provides a new direction that can substantially simplify the metrological protocol. We derive general conditions under which subsystem stabilizer codes achieve the Heisenberg limit and show that, for broad classes of noise, this can be realized by syndrome-free protocols using at most a single ancilla qubit. Furthermore, we extend this framework to dynamical error correction and show that Floquet codes can protect time-dependent metrological signals in reaching the Heisenberg limit.

About the speaker

Qiushi Liu is a postdoctoral researcher at Perimeter Institute for Theoretical Physics. He earned his PhD in computer science from the University of Hong Kong, supervised by Prof. Yuxiang Yang and Prof. Giulio Chiribella. Prior to that, he obtained a master in physics from ETH Zurich, and bachelor in physics from Peking University. His research interests include quantum metrology, quantum error correction and quantum foundations.

Title: Direct Likelihood Approach for Interval-Censored Competing Risks with Missing Failure Causes

Time: 10:30am 

Venue: Room 301, Run Run Shaw Building

Speaker(s): Professor Liming Xiang

Remark(s): 

Abstract

Interval-censored competing risks data with unknown causes of failure frequently appear in clinical studies, yet traditional two-stage estimation methods often suffer from high computational costs and efficiency loss. This talk introduces a direct likelihood approach under a mixture model framework to address these challenges. By incorporating competing risks and missing event types into a single likelihood function, the proposed method utilizes sieve maximum likelihood estimation to streamline computation and enhance estimation efficiency. We establish the consistency and asymptotic normality of the resulting estimators, demonstrate the method’s finite-sample performance through comprehensive simulations, and illustrate its practical utility using data from an Alzheimer’s disease study.

About the speaker

Interval-censored competing risks data with unknown causes of failure frequently appear in clinical studies, yet traditional two-stage estimation methods often suffer from high computational costs and efficiency loss. This talk introduces a direct likelihood approach under a mixture model framework to address these challenges. By incorporating competing risks and missing event types into a single likelihood function, the proposed method utilizes sieve maximum likelihood estimation to streamline computation and enhance estimation efficiency. We establish the consistency and asymptotic normality of the resulting estimators, demonstrate the method’s finite-sample performance through comprehensive simulations, and illustrate its practical utility using data from an Alzheimer’s disease study.

Title: Multi-modal AI for Biomedical Image Analysis and Visualisation

Time: 10:30am 

Venue: Innovation Wing Two, G/F, Run Run Shaw Building, HKU

Speaker(s): Prof. Jinman Kim

Remark(s): 

Abstract

Medical imaging plays a pivotal role in patient management in modern healthcare, with most patients who are treated in hospitals undergoing imaging procedures. These technologies can visualise anatomy and function in virtually every organ system in the body in intricate detail. There are numerous medical imaging modalities available; they vary in complexity and sophistication, from plain digital chest X-rays to simultaneous functional and anatomical imaging with positron emission tomography (PET) and computed tomography (CT) imaging (PET-CT). The challenge now is how to maximize the extraction of meaningful information from the images and present meaningful information to the users. There needs to be strategies to harness knowledge from vast image datasets and complementary sources like image sequences, text reports, and genomics. Fortunately, the era of artificial intelligence (AI) is fuelling the growth of smart decision support and analysis tools for medical image analysis. Despite rapid advancements in integrating AI algorithms into clinical decision support systems, we are still in the nascent stages of the AI revolution in medical imaging. This talk will present our research on multi-modal AI to integrate imaging and complementary data for disease modelling, analysis and visualization, aimed at improving the understanding, in an intuitive way.

About the speaker

"Jinman Kim is a Professor of Computer Science at the University of Sydney. He received his PhD from the University of Sydney in 2006 and was an Australian Research Council (ARC) Postdoctoral Research Fellow at the University of Sydney and then a Marie Curie Senior Research Fellow at the University of Geneva prior to joining the University of Sydney in 2013 as a faculty member. In 2024, he was a visiting professor at the Centre for Informatics at the University of Geneva, Switzerland. He is currently an ARC industry fellow, closely collaborating with his industry partner, Royal Prince Alfred Hospital, to conduct translational research. He is also the research director for Nepean AI research group at the Nepean Hospital. 

Prof Kim leads the Biomedical Data Analysis and Visualisation (BDAV) Lab at the School of Computer Science, pioneering research on the intersection of multimodal AI with biomedical data. His research focuses on includes biomedical visual-language representations, image-omics, multi-modal data processing, and biomedical mixed reality technologies. He has produced a number of publications in this field and received multiple competitive grants and scientific recognitions."