Title An Empirical Bayes Approach to Controlling the False Discovery Exceedance

Abstract In large-scale multiple hypothesis testing problems, the false discovery exceedance (FDX) provides a desirable alternative to the widely used false discovery rate (FDR) when the false discovery proportion (FDP) is highly variable. We develop an empirical Bayes approach to control the FDX. We show that, for independent hypotheses from a two-group model and dependent hypotheses from a Gaussian model fulfilling the exchangeability condition, an oracle decision rule based on ranking and thresholding the local false discovery rate (lfdr) is optimal in the sense that the power is maximized subject to the FDX constraint. We propose a data-driven FDX procedure that uses carefully designed computational shortcuts to emulate the oracle rule. We investigate the empirical performance of the proposed method using both simulated and real data and study the merits of FDX control through an application for identifying abnormal stock trading strategies. This is a joint work with Luella Fu, Alessio Saretto, and Wenguang Sun.

Short Bio. Pallavi Basu is an Assistant Professor of Operations Management (Statistics) at the Indian School of Business. Her research interests include applications of statistics in finance, marketing, operations, and other disciplines, high-dimensional statistical inference, large-scale multiple testing, and causal inference. Her methodological contributions have been published in top statistics journals as well as applied journals. She was a postdoctoral fellow at the Department of Statistics and Operations Research at Tel Aviv University, Israel. She obtained her Ph.D. in Business Administration from the Department of Data Sciences and Operations at USC Marshall School of Business, Los Angeles, USA.

Title Data-Pooling Reinforcement Learning for Personalized Healthcare Intervention

Abstract Motivated by the emerging needs of personalized intervention in many healthcare applications, we consider a class of multi-stage decision-making problems formulated by Markov Decision Process (MDP) with unknown model parameters. To deal with the pervasive issue of small sample size in personalized intervention, we develop a novel data-pooling estimator and the corresponding reinforcement learning (RL) algorithm based on a general perturbed value iteration framework. The model-free nature of our algorithm has the built-in benefits of being robust to model-misspecification and only requiring data-sharing via aggregate statistics. We establish a theoretical performance guarantee for our algorithm and show that it achieves strictly smaller regret bound than RL algorithms that use only target data or other data-pooling estimators. Furthermore, we substantiate our theoretical development with numerical results in a specific healthcare context: post-discharge follow-up intervention to prevent unplanned readmissions. We demonstrate the empirically better performance of our algorithm over a number of competitive benchmarks using a real hospital dataset. The talk is based on a joint work with Pengyi Shi and Shanwen Pu.

Short Bio. Xinyun Chen is currently an Assistant Professor in the School of Data Science at The Chinese University of Hong Kong, Shenzhen. She received her Ph.D in Operations Research from Columbia University in 2014. Her research interests include applied probability, stochastic simulation, queueing theory and reinforcement learning. She has published papers in journals and conferences including Annals of Applied Probability, Mathematics of Operations Research, Stochastic Systems and ICLR and currently serves on the editorial board of the Applied Probability Trust.

Title Non-reversible guided Metropolis kernel

Abstract. We construct a class of non-reversible Metropolis kernels as a multivariate extension of the guided-walk kernel proposed by Gustafson 1998. The main idea of our method is to introduce a projection that maps a state space to a totally ordered group. By using Haar measure, we construct a novel Markov kernel termed Haar-mixture kernel, which is of interest in its own right. This is achieved by inducing a topological structure to the totally ordered group. Our proposed method, the Delta-guided Metropolis–Haar kernel, is constructed by using the Haar-mixture kernel as a proposal kernel. The proposed non-reversible kernel is at least 10 times better than the random-walk Metropolis kernel and Hamiltonian Monte Carlo kernel for the logistic regression and a discretely observed stochastic process in terms of effective sample size per second.

* Talk is held physically at NUS seminar room S17-05-11.

Title Information-Theoretic Characterization of the Generalization Error for Iterative Semi-Supervised Learning

Abstract Using information-theoretic principles, we consider the generalization error (gen-error) of iterative semi-supervised learning (SSL) algorithms that iteratively generate pseudo-labels for a large amount of unlabelled data to progressively refine the model parameters. In contrast to most previous works that bound the gen-error, we provide an exact expression for the gen-error and particularize it to the binary Gaussian mixture model. Our theoretical results suggest that when the class conditional variances are not too large, the gen-error decreases with the number of iterations, but quickly saturates. On the flip side, if the class conditional variances (and so amount of overlap between the classes) are large, the gen- error increases with the number of iterations. To mitigate this undesirable effect, we show that regularization can reduce the gen-error. The theoretical results are corroborated by extensive experiments on the MNIST and CIFAR datasets in which we notice that for easy- to-distinguish classes, the gen-error improves after several pseudo-labelling iterations, but saturates afterwards, and for more difficult-to-distinguish classes, regularization improves the generalization performance. 

Short Bio. Haiyun He is currently a Ph.D. student in the Department of Electrical and Computer Engineering (ECE) at the National University of Singapore (NUS). She has successfully defended her thesis in August and will receive the degree at end September. From Sep 2017 to Jul 2018, she was first a Research Assistant in ECE at NUS. She received the B.E. degree in Beihang University (BUAA) in 2016 and the M.Sc. (Electrical Engineering) degree in ECE from NUS in 2017. Her research interests include information theory, statistical learning and their applications.

Title Scalable and Reliable Inference for Probabilistic Modeling

Abstract Probabilistic modeling provides a principled framework for learning from data, with the key advantage of offering rigorous solutions for uncertainty quantification. In the era of big and complex data, there is an urgent need for new inference methods in probabilistic modeling to extract information from data effectively and efficiently. 

This talk will show how to do theoretically guaranteed scalable and reliable inference for modern machine learning. First, I will introduce a general and theoretically grounded framework to enable fast and asymptotically correct inference, with applications to Metropolis-Hastings and Gibbs sampling. Then, I will highlight the key challenges of probabilistic inference in deep learning, and present a novel approach for fast posterior inference of neural network weights. This method has achieved state-of-the-art results and has been regarded as one of the benchmarks in deep probabilistic modeling. 

Short Bio. Ruqi Zhang is an Assistant Professor in the Department of Computer Science at Purdue University. Before that, she was a postdoctoral fellow at the Institute for Foundations of Machine Learning at UT Austin. She obtained her Ph.D. in Statistics at Cornell University. Her research focuses on probabilistic methods for machine learning and data science. She currently works on developing scalable and reliable Bayesian methods with theoretical guarantees and their applications with modern model architectures, such as deep neural networks, on real-world big data. Her work has been published in top machine learning venues such as NeurIPS, ICML, ICLR, and AISTATS, and has been recognized through several Oral and Spotlight Awards.

Title Landscape modification meets spin systems: from torpid to rapid mixing and tunneling in the low-temperature regime

Abstract This talk centers around a technique that we call landscape modification. The core idea is that the Hamiltonian function is suitably modified in a way for rapid mixing while maintaining proximity with the original target distribution. We first present model-independent results that give rapid mixing and tunneling in the low-temperature regime. Building upon these results, we investigate the effect of landscape modification on some prototypical statistical physics models including the Ising model on various deterministic and random graphs as well as Derrida’s random energy model. This talk highlights a novel use of the geometry and structure of the landscape, in particular the global minimum value or its estimate, to the design of accelerated samplers or optimizers.

Short Bio.  Michael Choi is currently an Assistant Professor at the Department of Statistics and Data Science at National University of Singapore. He also holds a joint appointment at the Yale-NUS College and is affiliated with the Institute of Operations Research and Analytics (IORA). His research centers around the theory and applications of Markov chains and Markov processes, with a particular emphasis on the design and analysis of stochastic algorithms driven by Markov chains such as Markov chain Monte Carlo (MCMC). He is broadly interested in related areas such as statistical physics, applied probability, stochastic optimization and Bayesian statistics. His research has been recognized by an invited plenary lecture at the Workshop on non-reversible Markovian Monte Carlo at Lorentz Center, Leiden.

* Talk is held in face-to-face format at room S17-0511.

Title A Neural Network Approach for Homogenization of Multiscale Problems

Abstract We propose a neural network-based approach to the homogenization of multiscale problems. The proposed method uses a derivative-free formulation of a training loss, which incorporates Brownian walkers to find the macroscopic description of a multiscale PDE solution. Compared with other network-based approaches for multiscale problems, the proposed method is free from the design of hand-crafted neural network architecture and the cell problem to calculate the homogenization coefficient. The exploration neighborhood of the Brownian walkers affects the overall learning trajectory. We determine the bounds of micro- and macro-time steps that capture the local heterogeneous and global homogeneous solution behaviors, respectively, through a neural network. The bounds imply that the computational cost of the proposed method is independent of the microscale periodic structure for the standard periodic problems. We validate the efficiency and robustness of the proposed method through a suite of linear and nonlinear multiscale problems with periodic and random field coefficients.

Short Bio. PhD in Mathematics, University of Texas at Austin, 2013 (advisor: Bjorn Engquist) Postdoc, Courant Institute, 2013-2017 (Mentor: Andrew Majda; same as Prof Xin Tong’s) Research scientist, Lawrence Berkeley Lab, 2017-2018 Assistant Professor of Mathematics, Dartmouth College, 2018-current

Title Biological sequence analysis via deep learning

Abstract Known as the blueprint of life, the genomic sequence contains instructions for controlling a species’ growth, development, survival, and reproduction. From the perspective of Computer Science, the primary structure of the genomic sequences is simply a string/sequence defined on a small alphabet {A, C, G, T}. Different parts of the sequence have different biological functions and thus comprise the “language of life”.

Next-generation sequencing (NGS) technologies, which produce vast amount of sequencing data for various life forms, have provided tremendous information for tackling grand challenges in various fields. NGS enables us to obtain a large amount of data to understand more about the language of life. In this talk, I will focus on my research on using deep learning models to make sense out of the BIG microbial sequencing data obtained from host-associated (such as human gut, skin) or environmental samples (such as water, soil). By using protein clusters as the tokens (“words”), we model viral genomes as a protein-based language and thus use modern language model Transformer for feature learning. In addition, to address the issues of limited labeled biological data, we use graph convolutional network (GCN), a semi-supervised learning model for leveraging both labeled and unlabeled data for feature learning in several tasks. These research projects rendered us with deeper understanding of the advantages and limitations of using deep learning models for biological sequence classification. I will share some of our observations along this line.

Short Bio. Yanni Sun is an associate professor in Electronic Engineering at City University of Hong Kong. Before she relocated to Hong Kong, she was an Associate Professor in the Department of Computer Science and Engineering at Michigan State University, USA. She received the B.S. and M.S. degrees from Xi’an JiaoTong University (China), both in Computer Science. She received the Ph.D. degree in Computer Science and Engineering from Washington University in Saint Louis, USA. She works in bioinformatics and computational biology. Her recent research interests include sequence analysis, application of machine learning and data mining models in analyzing next-generation sequencing data, metagenomics, protein domain annotation, and noncoding RNA annotation. She was a recipient of NSF CAREER Award in 2010.

* The talk is held at 2 pm on 22 Dec 2022 at S17-05-12.