About me

I am currently an Assistant Professor HES in Business Analytics at the Haute École de Gestion de Genève (HEG Geneva), part of the University of Applied Sciences, Western Switzerland. I teach various topics around data science, information systems and business analytics, conduct research in artificial intelligence, theoretical and applied machine learning and statistics, as well as write open source software. I am also pleased to offer my expertise for consulting and implementation services in public-private collaborations.

Short Bio

In 2010, I earned a BSc in Physics, followed by an MSc in Physics/Financial Engineering in 2012 at the Swiss Federal Institute of Technology in Lausanne (EPFL). Subsequently, I obtained a PhD in Statistics from HEC Lausanne in 2016, focusing on generalized additive models for conditional dependence structures among random variables.

Following my doctoral studies, in 2017, I then received a Swiss National Science Foundation fellowship and moved to New York to join the Statistics Department of Columbia University as a postdoctoral researcher. Within a year, in 2018, I was promoted to Assistant Professor to teach as well as keep pursuing my research in artificial intelligence, theoretical and applied machine learning and statistics, contributing to top-tier journals and conferences.

In January 2022, I then left academia for the tech industry, joining Meta (Facebook) as a senior Research Data Scientist. Partnering with multiple infrastructure teams, I developed data-driven solutions for large scale video infrastructure challenges.

Notably, I also had two experiences as a Quant in the finance industry:

  • Between my MSc and PhD (2012), I have interned at Swissquote Bank Ltd, implementing methods to blend subjective investor’s views in quantitative portfolio allocations.
  • Between my PhD and Postdoc (2016), I have worked at swissQuant Group AG to develop risk model for large portfolio of financial derivatives on energy commodities.

In the Fall of 2023, I returned to Geneva to join the University of Applied Sciences, Western Switzerland, as an Assistant Professor HES in Business Analytics.

Interests
  • Methodology Artificial Intelligence, Machine Learning, Computational Statistics, Multivariate Analysis, Time-Series
  • Programming
    Python/SQL/C++/R/MATLAB
  • Applications
    Financial Econometrics, Business Analytics, Operations Research
Education

  • PhD in Statistics, 2016

    HEC Lausanne

  • MSc in Physics, 2012

    Swiss Federal Institute of Technology, Lausanne

  • BSc in Physics, 2010

    Swiss Federal Institute of Technology, Lausanne

Copula Covariates Deep learning Financial derivatives Financial econometrics Financial mathematics Foreign exchange returns Generative modeling Intraday Machine learning Nonparametric Python Quantile regression R Semiparametric Smoothing Splines Statistics Vine copula vinecopulib

Publications

Quickly discover relevant content by filtering publications.
Damien Ackerer, Natasha Tagasovska, Thibault Vatter (2020). Deep Smoothing of the Implied Volatility Surface. Advances in Neural Information Processing Systems (NeurIPS 2020).

Cite URL

Natasha Tagasovska, Valérie Chavez-Demoulin, Thibault Vatter (2020). Distinguishing Cause from Effect Using Quantiles: Bivariate Quantile Causal Discovery. Proceedings of the 37th International Conference on Machine Learning (ICML2020).

Cite URL

Natasha Tagasovska, Damien Ackerer, Thibault Vatter (2019). Copulas as High-Dimensional Generative Models: Vine Copula Autoencoders. Advances in Neural Information Processing Systems 32 (NeurIPS 2019).

Cite URL

V Kulkarni, Natasha Tagasovska, Thibault Vatter, Benoit Garbinato (2018). Generative Models for Simulating Mobility Trajectories. NeurIPS 2018 Workshop on Control and decision making in Spatiotemporal domain.

PDF Cite

Thibault Vatter, Thomas Nagler (2018). Generalized Additive Models for Pair-Copula Constructions. Journal of Computational and Graphical Statistics.

Cite URL

See all publications

Software

*
All R C++ Python MATLAB
pyvinecopulib
High Performance Algorithms for Vine Copula Modeling in Python
Last updated on Mar 4, 2021
Cite
rvinecopulib
High Performance Algorithms for Vine Copula Modeling in R
Last updated on Mar 4, 2021
Cite
eecop
An R Package to Solve Estimating Equations with Copulas
Last updated on Mar 4, 2021
Cite
gamCopula
Generalized Additive Models for Bivariate Conditional Dependence Structures and Vine Copulas
Last updated on Mar 4, 2021
Cite
kde1d
Univariate Kernel Density Estimation
Last updated on Mar 4, 2021
Cite
VineCopula
Statistical Inference of Vine Copulas
Last updated on Mar 4, 2021
Cite
vinecopulib
High Performance Algorithms for Vine Copula Modeling
Last updated on Mar 4, 2021
Cite
intradaySST
Tools to Extract Trends and Seasonality from Intraday Financial Data
Last updated on Mar 4, 2021
Cite
mgpancestry
Python Classes to Scrap the Mathematical Genealogy Project for a Scholar’s Ancestry
Last updated on Mar 4, 2021
Cite

Talks

2020

  • JSM 2020 (online)
  • ICML 2020 (online)
  • Community of Practice for Compliance: workshop (online)

2019

  • CMStatistics 2019 (London)
  • European Meeting of Statisticians (Palermo)
  • Vine Copulas and their Applications (Munich)
  • UNIL (Lausanne)

2018

  • MIT (Boston)
  • TUM (Munich)
  • Columbia (New York)

2017

  • CMStatistics 2017 (London)
  • Chatham Financial (Kennett Square)
  • Columbia University (New York)

2016

  • TUM (Munich)
  • UNIL (Lausanne)

2015

  • swissQuant Group AG (Zurich)
  • UNINE (Neuchâtel)

2014

  • TUM (Munich)
  • CMStatistics 2014 (Pisa)
  • UNIL (Lausanne)

2013

  • CFE 2013 (London)
  • EPFL (Lausanne)
  • UNISTRA (Strasbourg)
  • UC Berkeley (Berkeley)

Teaching

2023

  • Introduction to Data Science (HEG Geneva)
  • Information Systems (HEG Geneva)
  • Information Technologies and Web Analytics (HEG Geneva)
  • Data Science for Business Analytics (HEC Lausanne)

2021

  • Computational Statistics and Introduction to Data science (Columbia STAT GU5206)
  • Data Science for Business Analytics (HEC Lausanne)

2020

  • Computational Statistics and Introduction to Data science (Columbia STAT GU5206)
  • Data Science for Business Analytics (HEC Lausanne)

2019

  • Computational Statistics and Introduction to Data science (Columbia STAT GU5206)
  • Statistical Inference (Columbia STAT GU4204)
  • Data Science for Business Analytics (HEC Lausanne)

2018

  • Statistical Inference (Columbia STAT GU4204 / GR5204)
  • Time series analysis (Columbia STAT GU4221 / GR5221)
  • Data Science for Business Analytics (HEC Lausanne)

2017

  • Math Methods for Political Science (Columbia POLS GU4700)

Consulting

I am pleased to offer my expertise for consulting and implementation services in public-private collaborations. With a robust background and significant experience in both academia and industry, I am uniquely positioned to bridge the gap between cutting-edge research and practical application in the business world. Services are available through direct contracting or via Innosuisse grants (see below). For inquiries, please contact me.

Key Skills and Expertise

Data Science and Analytics

  • Data Wrangling, Data Visualization, Data Mining, Data Exploration with Python, R, SQL
  • Predictive Modeling, Machine Learning and Artificial Intelligence, Statistical Inference, Experimental Design
  • Big Data Technologies, Data Quality Control

Machine Learning and Artificial Intelligence

  • Supervised and Unsupervised Learning, Feature Engineering, Model Selection and Tuning
  • Deep Learning, Neural Networks, Natural Language Processing, Computer Vision
  • Model Validation, Model Deployment, MLOps
  • Computational Statistics, Time Series Analysis, Multivariate Analysis, Dependence Modeling

Software and Data Engineering

  • Data Engineering, DataOps (e.g., Airflow and Kafka), ETL
  • Software Development, System Architecture, Database Management
  • Cloud Computing (e.g., AWS), Big Data Platforms, Scalability Solutions
  • Code Optimization, Technical Documentation
  • Continuous Integration/Continuous Deployment (CI/CD), Automation

Quantitative Finance and Risk Management

  • Quantitative Finance, Portfolio Optimization, Performance Analysis
  • Quantitative Risk Management, Option Pricing Models, Fixed Income Analysis
  • Credit Risk, Financial Econometrics

Business and Technology Leadership

  • Business Analytics, Analytics and AI Strategy, Project Coordination
  • Cross-Functional Collaboration, Stakeholder Engagement
  • Team Leadership, Mentoring, Training

Innosuisse Grants

Innovation Projects with Implementation Partners

Engage in groundbreaking innovation projects by partnering with a research institution. As an industry partner, whether from the private or public sector, you have the opportunity to apply jointly for financial support. These projects are ideal if they hold the promise of commercial success or societal benefit. I can guide you through the process of application, development, and implementation, ensuring your innovative ideas reach their full potential.

Vouchers for Preliminary Studies: Innovation Cheques

For SMEs or organizations with fewer than 250 full-time equivalents, there is an opportunity to receive an Innovation Cheque worth up to 15,000 Swiss francs. This funding is designed to evaluate the feasibility of innovative concepts. My role involves aiding in preliminary studies, idea development and evaluation, and analyzing both innovation and market potentials.

Contact