Joint Econometrics and Statistics Seminar Series

This event will take place in the Leverhulme Library (COL 6.14).

Title: Confounder Selection via Iterative Graph Expansion

Abstract: Confounder selection, namely choosing a set of covariates to control for confounding between a treatment and an outcome, is arguably the most important step in the design of observational studies. Previous methods, such as Pearl’s celebrated back-door criterion, typically require pre-specifying a causal graph, which can often be difficult in practice. We propose an interactive procedure for confounder selection that does not require pre-specifying the graph or the set of observed variables. This procedure iteratively expands the causal graph by finding what we call “primary adjustment sets” for a pair of possibly confounded variables. This can be viewed as inverting a sequence of latent projections of the underlying causal graph. Structural information in the form of primary adjustment sets is elicited from the user, bit by bit, until either a set of covariates are found to control for confounding or it can be determined that no such set exists. Other information, such as the causal relations between confounders, is not required by the procedure. We show that if the user correctly specifies the primary adjustment sets in every step, our procedure is both sound and complete.

Bio: Qingyuan is a University Associate Professor of Statistics in the Statistical Laboratory, Department of Pure Mathematics and Mathematical Statistics (DPMMS) at University of Cambridge, a Fellow of the Corpus Christi College, and an Associate Faculty of the Cambridge Centre for AI in Medicine (CCAIM).

He is interested in improving the general quality and appraisal of statistical research, including new methodology and a better understanding of causal inference, novel study designs, sensitivity analysis, multiple testing, and selective inference.

This event will take place in the Data Science Institute (COL.1.06).

Title: Consistent Causal Inference for High-Dimensional Time Series

Abstract: A methodology for high dimensional causal inference in a time series context is introduced. It is assumed that there is a monotonic transformation of the data such that the dynamics of the transformed variables are described by a Gaussian vector autoregressive process. This is tantamount to assuming that the dynamics are captured by a Gaussian copula. No knowledge or estimation of the marginal distribution of the data is required. The procedure consistently identifies the parameters that describe the dynamics of the process and the conditional causal relations among the possibly high dimensional variables under sparsity conditions. The methodology allows us to identify such causal relations in the form of a directed acyclic graph. As illustrative applications we consider the impact of supply side oil shocks on the economy, and the causal relations between aggregated variables constructed from the limit order book on four stock constituents of the S&P500.

Bio: Alessio has been professor of economics at Royal Holloway, University of London, since 2012.  I previously held a University lectureship at the University of Cambridge, where I completed my PhD in 2002. His research interests lie in financial economics and econometrics, asset pricing, market microstructures and manipulation, forecasting methods, high dimensional estimation and machine learning.

This event will take place in the Data Science Institute (COL.1.06).

Title: Bayesian Optimal Experimental Design

Abstract: In order to use machine learning in a domain without pre-existing data, we must first gather data by conducting experiments. This presents an opportunity, because we can design experiments carefully to ensure we gather the most informative data. As we gather a little data, we can use this to guide the design of future experiments in an adaptive manner. Bayesian Optimal Experiment Design (BOED) is a mathematical framework that precisely defines the optimal experiment in a pipeline like this. Historically, though, BOED has been computationally too challenging to use in practice. In this talk, I discuss recent computational advances in the field that utilise variational inference, gradient-based optimisation and policy learning to overcome many of these computational bottlenecks. I will also touch on applications such as adaptive survey design and force-field learning in AI for science.

Bio: I am a senior researcher at Microsoft Research AI4Science where I work on machine learning methods for chemistry with Frank Noé and Jan Hermann, focusing on QMC. I also have a strong interest in Bayesian experimental design and active learning. I am driven by the desire to understand how machine learning can help us to solve critical problems in the sciences and to build new, sustainable technology. Previously, I did my PhD in Statistical Machine Learning at the University of Oxford, supervised by Yee Whye Teh and Tom Rainforth in the Computational Stats and Machine Learning Group in the Department of Statistics. Before starting my PhD, I studied mathematics at Cambridge where my Director of Studies was Julia Gog.

Title: The impact of job stability on monetary poverty in Italy: causal small area estimation

Abstract: Job stability refers to the security and predictability of employment, including factors such as long-term contracts, adequate wages, social security benefits, and access to training and career development opportunities. Stable employment can play a crucial role in reducing poverty, as it provides individuals and households with a stable income as well as improves their overall and subjective economic well-being. In this work, we leverage the EU-SILC survey and census data to assess the causal effect of job stability on monetary poverty across provinces in Italy. To this end, we propose a causal small area estimation (CSAE) framework for heterogeneous treatment effect estimation in which only a negligible fraction of outcomes is actually observed at the provincial level. Our estimators are more stable than the classical causal inference tools as they borrow strength from the other sources of data at the expense of additional modelling assumptions. On top of that, our new methodology proves to be successful in recovering provincial heterogeneity of the effect of job stability across six regions in Italy.

Bio: I am a Lecturer (Assistant Professor) in the School of Mathematics at the University of Bristol, affiliated with the Institute for Statistical Science. Prior to joining the University of Bristol, I did postdocs at the University of California, Berkeley, the University of Toronto, and the University of Cambridge, where I worked with Mark van der Laan, Dehan Kong, and Qingyuan Zhao, respectively. I completed my PhD in the University of Geneva, where I was advised by Stefan Sperlich and co-advised by María José Lombardía. My research interest lies at the intersection of survey methodology, causal inference and machine learning. During my PhD I worked on theoretical aspects of simultaneous, post-selection and computational inference with some applications in economics and social sciences. Afterwards, I broadened my research agenda by trying to solve some open problems in causal inference and merging machine learning with survey sampling methodology.

This event will take place in the Data Science Institute (COL.1.06).

Title: Identifiability and Consistent Estimation for Gaussian Chain Graph Model

Abstract: The chain graph model admits both undirected and directed edges in one graph, where symmetric conditional dependencies are encoded via undirected edges and asymmetric causal relations are encoded via directed edges. Though frequently encountered in practice, the chain graph model has been largely under investigated in the literature, possibly due to the lack of identifiability conditions between undirected and directed edges. In this paper, we first establish a set of novel identifiability conditions for the Gaussian chain graph model, exploiting a low rank plus sparse decomposition of the precision matrix. Further, an efficient learning algorithm is built upon the identifiability conditions to fully recover the chain graph structure. Theoretical analysis on the proposed method is conducted, assuring its asymptotic consistency in recovering the exact chain graph structure. The advantage of the proposed method is also supported by numerical experiments on both simulated examples and a real application on the Standard & Poor 500 index data.

Bio: Haoran Zhang is currently an Assistant Professor in the Department of Statistics and Data Science at Southern University of Science and Technology (SUSTech), where he also serves as a doctoral supervisor. He obtained his Ph.D. in Statistics from the Shanghai Center for Mathematical Sciences at Fudan University in 2021 supervised by Prof. Zhiliang Ying, after completing his undergraduate studies at the School of Mathematical Sciences, Fudan University in 2016. From 2021 to 2023, he conducted postdoctoral research at The Chinese University of Hong Kong and City University of Hong Kong, and from 2017 to 2019, he was a visiting scholar at the Department of Statistics at Columbia University. His research interests include network analysis, graphical models, causal inference, and psychometrics. He has published papers in prestigious journals such as the Journal of the American Statistical Association, Journal of Machine Learning Research, and Psychometrika.

This event will take place in the Data Science Institute (COL.1.06).

Title: Regression Analysis of Neuroimaging Data Leveraging Deep Neural Networks

Abstract: The complex interplay between neuroimaging data and variables of interest poses significant challenges for conventional regression models. These challenges are due to the ultra-high dimensionality, varying levels of noise, and limited sample sizes inherent to this type of data. In this talk, I will introduce a series of regression models specifically designed for neuroimaging data analysis, utilizing Deep Neural Networks (DNN) to enable more accurate statistical inference. Unlike traditional approaches, our innovative methods offer enhanced flexibility in capturing intricate patterns in brain activity, accommodating the heterogeneity in noise levels and spatial dependencies across different brain regions. I will delve into parameter estimation, inference procedures, and the theoretical underpinnings of our advanced models, ultimately demonstrating their superior performance over existing methodologies through rigorous simulations and real-world neuroimaging case studies.

Bio: TBC

This event will take place in the Data Science Institute (COL.1.06).

Title: Estimation of Grouped Time-Varying Network Vector Autoregression Models

Abstract: This paper introduces a flexible time-varying network vector autoregressive model framework for large-scale time series. A latent group structure is imposed on the heterogeneous and node-specific time-varying momentum and network spillover effects so that the number of unknown time-varying coefficients to be estimated can be reduced considerably. A classic agglomerative clustering algorithm with nonparametrically estimated distance matrix is combined with a ratio criterion to consistently estimate the latent group number and membership. A post-grouping local linear smoothing method is proposed to estimate the group-specific time-varying momentum and network effects, substantially improving the convergence rates of the preliminary estimates which ignore the latent structure. We further modify the methodology and theory to allow for structural breaks in either the group membership, group number or group-specific coefficient functions. Numerical studies including Monte-Carlo simulation and an empirical application are presented to examine the finite-sample performance of the developed model and methodology.

Bio: I joined the Department of Mathematics at University of York in March 2013 and was promoted to professor in January 2018. Prior to that I worked as a Research Fellow and Senior Research Fellow at Monash University in Australia (2011-2013) and a Research Associate at University of Adelaide in Australia (2008-2010). More details can be found in my personal website here.