In 2016, I obtained a PhD in Statistics from HEC Lausanne under the supervision of Valérie Chavez-Demoulin. As a recipient of a fellowship from the Swiss National Science Foundation, I then conducted my postdoctoral research in the group of Richard Davis at Columbia University. I am now an assistant professor in the statistics department at Columbia University.
I am currently working on diverse methodological and applied projects, ranging from solving statistical learning problems using copulas to modeling time-series of counts semiparametrically, and from building high-dimensional risk models for thousands of assets to developing new causal discovery algorithms.
PhD in Statistics, 2016
MSc in Physics, 2012
BSc in Physics, 2010
Vatter, T. and Tagasovska, N. (2018)
Nonparametric Vines as Generative Models
Vatter, T., Ackerer, D. and Nagler, T. (2018)
High-Dimensional Pair-Copula Constructions with Financial Applications
Davis, R. and Vatter, T. (2018)
Zero-Inflated Time Series of Counts with Shape Constraints
De Treville, S., Hoffstetter, J. and Vatter, T. (2018)
Using Point-of-Sale Data To Improve Shelf Replenishment Performance
Nagler T., Vatter, T. (2018)
Solving Estimating Equations with Copulas
Tagasovska, N., Vatter, T. and Chavez-Demoulin, V. (2018)
Nonparametric Quantile-based Causal Discovery
Vatter, T. and Nagler, T. (2018)
Generalized Additive Models for Pair-Copula Constructions
Journal of Computational and Graphical Statistics, to appear
Ackerer, D. and Vatter, T. (2017)
Dependent Defaults and Losses with Factor Copula Models
Dependence Modeling, 5:375–399
Vatter, T. and Chavez-Demoulin, V. (2015)
Generalized Additive Models for Conditional Dependence Structures
Journal of Multivariate Analysis, 141:147-167
Vatter, T., Wu, H.-T., Chavez-Demoulin, V. and Bin, Y. (2015)
Non-Parametric Estimation of Intraday Spot Volatility: Disentangling Instantaneous Trend and Seasonality
Econometrics, 3:864-887
Vatter, T. (2016)
Generalized Additive Modeling for Multivariate Distributions
PhD thesis, Faculty of Business and Economics, University of Lausanne
High-performance C++ library for vine copula modeling based on Boost and Eigen
R interface to the vinecopulib C++ library
Efficient implementation of univariate local polynomial kernel density estimators that can handle bounded and discrete data.
Generalized additive models for bivariate conditional dependence structures and vine copulas
Statistical inference of vine copulas
Copula-based causal discovery and directed acyclic graphs, accompanies the paper Nonparametric Quantile-based Causal Discovery
Tools to model multivariate discrete mixture distributions
Tools to solve estimating equations with copulas, accompanies the paper Solving Estimating Equations with Copulas.
Scraping the Mathematical Genealogy Project for a scholar’s ancestry
Functions to clean intraday FX prices and apply the Synchrosqueezing transform to extract the trend and seasonality from the returns volatilities, accompanies the paper by Vatter, T. and al. (2015)