Research

Prof. Valérie Chavez orientates her research towards many areas: Extreme Value Theory, Dependence Modelling, Quantitative Risk Management, Operational Risk, Graphical Causal Models, Learning Analytics as well as Quantitative Methods for Operations Management.

She is specialising in statistical methods for quantitative risk management in general, and the statistical modeling of extreme events in particular.

Her more recent methodological work concerns conditional dependence structures modeling, non-parametric Bayesian models, dynamic Extreme Value Theory models and extremes for non-stationary time series. 

A significant part of Prof. Chavez's research benefits from her (inter)national scientific collaborations and the invitations for conferences and seminars around the world including the Erasmus Universiteit of Rotterdam, Columbia University, EPFL, CERN, ETH Zurich, University of Geneva, University of California Berkeley, McGill University in Montreal, University of Strasbourg, LSCE, CNRS, University of Edinburgh and the University of Waterloo.

She is also member of the RiskLab at ETH of Zurich and is an elected member of ISI (The International Statistical Institute).

Mathematical Logic in general and more precisely, Descriptive Set Theory and its relations to Theoretical Computer Science. Topological complexities and games of infinite length.

Borel mappings via games and representation theorems
We characterize some classes of Borel functions over the Baire space as strategies in suitable games. This is a way towards both obtaining representation theorems and elaborating a fine classification of Borel functions. Representation theorems come as a representation of some classes of functions very different from their definitions. For instance, a cornerstone for this type of result is the Baire Grand Theorem which states that the following are equivalent for any function f on a Polish space:

- f is the pointwise limit of countably many continuous functions;
- on every non-empty closed subset f admits a point of continuity.

Continuous Reductions on Quasi-Polish spaces
Quasi-Polish spaces is a novel unifying theory due to Matthew de Brecht. It brings together topological structures that were previously unrelated. It connects closely topology in mathematical analysis -- which is usually Hausdorff (T_2) -- to topology in computer science -- which is rather Kolmogorov (T_0) — by offering the Polish spaces as well as the omega-algebraic or omega-continuous domains a common roof. Quasi-Polish spaces are derived from Polish spaces -- which are separable completely metrizable topological spaces -- by simply relaxing the symmetry condition in the definition of a metric.

We propose to design and make use of game theoretical tools to study the reductions between these sets and explore the underlying ordering as well as the natural hierarchies that would arise. We intend to do this in a similar manner as the way we studied the Wadge hierarchy of Borel subsets of the Cantor space.

Quotients of Projective Fraïssé Limits
The idea of studying infinite structures via approximation by finite structures is a well rooted concept in mathematics. In particular, the Fraïssé limit is an extensively studied tool in many areas of mathematics.
In 2006 T. Irwin and S. Solecki introduced the projective Fraïssé limit of topological structures. Many applications have since been found in continua theory and descriptive dynamics.
We propose to isolate and study the class of all compact metric spaces that are obtainable as a quotient of a projective Fraïssé limit by the interpretation of a binary relation symbol from the language. Our hope is to describe a natural way of obtaining such spaces.

The Wadge Hierarchy
Over a century ago, the modern theory of integration, based on measure theory induced a fundamental interest in the study of well-behaved subsets of the real line or the real plane. Topology, which developed about the same time yielded the mathematical framework for such a study. For instance, the σ-algebra generated by the open subsets proved to be central in measure theory, for the sets it defines bear all desired nice properties. The most refined classification of these sets is the so-called Wadge hierarchy whose study involves methods from (set theoretical) game theory

Topological Complexity, Games, Logic and Automata
We try to unravel the fine topological structure of omega-regular tree languages which are the infinitary languages of trees recognized by automata. In other words, we exhibit the Wadge hierarchy of non deterministic omega-tree automata.

University of Genoa, University of Turin, Paris Diderot University, Pierre and Marie Curie University, Warsaw University, RWTH Aachen

Treasurer of the Swiss Society for Logic and Philosophy of Science 

Member of the Swiss Mathematical Society 

During the past 20 years Prof. Ari-Pekka Hameri has performed over 100 supply chain audits in various industries, which include pulp & paper, electronics, textiles, food and beverage, metal industry, watch, medical instrument and some service sectors.

He has also studied workflows and document processes related to new product development and project management.

Prof. Hameri has published over 70 articles in international journals on production, project and supply chain management, and on industrial IT. 

His current research interest are related to patient flows in general hospitals, and on how internet of things can improve supply chain performance by transferring traditional logistic systems into supply chain platforms. He is also involved with research initiatives concerning supply chain security through the Cross-Border Research Association.

Prof. Hameri is currently working with multi-nationals on supply chain forecasting and supply chain security. Patient flow research is being done in collaboration of CHUV. He also has active research relationship with CERN.

Prof. Olivier Gallay focuses mainly on operations research and on the optimization of complex systems. His research axis encompass in particular transportation, logistics, and supply chain management. More generally, his research interests also include applied stochastic processes, non-linear dynamics, and agent-based models.

He currently works on the development of new optimization algorithms for the operational management of multi-modal transportation networks. Focusing on sustainable logistics, different transport modes are efficiently combined in order to provide home services at customer locations while reducing the ecological footprint associated with these activities. Another active project focuses on the analysis of imitation mechanisms within a population, in particular with respect to their influence on the diffusion of innovation. In that framework, agent-based models are built in order to understand the emergence of global collective patterns from the microscopic agent interactions.

Prof. Olivier Gallay works extensively with both various academical and industrial partners. Currently, he has active collaborations with Renault, and with a major European energy provider. Academical partnerships include presently EPFL, Karolinska Institutet, and the VTT Technical Research Centre of Finland. 

Prof. Virginie Lurkin’s research aims at developing innovative solutions for sustainable urban mobility and logistics, using multidisciplinary methodologies rooted in operations management, discrete choice models, and data-driven decision making.

Prof. Virginie Lurkin is currently supervising 4 doctoral theses. Three of these projects focus optimization problems in the fields of urban mobility and logistics. The fourth concerns the development of models of oligopolistic competition that integrate discrete-choice models. She is also working on pricing models derived from game theory (competitive and collaborative) that take into account individual consumer choices.

Prof. Virginie Lurkin is co-supervising a doctoral thesis on oligopolistic competition models that integrate discrete-choice models with Professor Michel Bierlaire from EPFL, and co-supervising two doctoral students on urban logistics optimization projects with Prof. Tom van Woensel from Eindhoven University of Technology.

Prof. Gautier Stauffer is using mathematical modelling and analysis, and (mixed) integer programming and combinatorial optimisation in particular, to address challenging decision and optimization problems in operations and supply chain management.

He is currently working on the following research projects:

- Leveraging cargo bikes for the design of sustainable urban parcel delivery networks (Groupe La Poste, France)

- Patient flow optimization in Intensive Care Units (Hôpital des armées de Toulon)

- Optimisation of the assembly of emergency kits (MSF Logistique)

- Fostering (horizontal) collaboration in supply chain management through collaborative game theory (CNRS, University of Rome)

- Forecasting and optimizing professional golfer’s performances using statistical inference and Markov Decision Processes
 

- Approximation algorithms fo some stochastic joint replenishment problems  using cargo bike for green delivery
 

Current industrial partners: Groupe La Poste, Hôpital des armées de Toulon, MSF Logistique

Current academic partners: Georgia Tech, University of Rome Tor Vergata, CNRS, Kedge Business School


 

Prof. Ann van Ackere is a Management Scientist. Her methodological focus includes System Dynamics, stochastic simulation, optimisation and game theory. Her main application areas are energy policies, queuing (service management) and health management.

Prof. Ackere currently works on a project looking at the co-evolution of electricity markets and their regulation, funded by the Swiss National Science Foundation.

Prof. Suzanne de Treville’s research is centered around how companies can use manufacturing to improve their profitability and innovation while reducing their carbon footprint. Her laboratory develops decision tools based on quantitative finance that allow decision makers to value the responsiveness that comes from compressing the supply chain. Other decision tools aid firms in reducing the manufacturing lead time: Local production is only responsive if the local plant is able to commit production after observing demand.

The tools developed by Prof. de Treville’s lab OpLab are being actively deployed by several governments (national and cantonal) to support local manufacturing. The lab is also developing competitive games that make the decision tools more intuitive and interactive for decision makers.

Prof. de Treville will serve as co-Editor-in-Chief of the Journal of Operations Management from 2018-2020, which is a Financial Times top 50 academic journal, after guest-editing a special issue of the same journal on competitive manufacturing in a high-cost economy. She has regularly served as visiting professor at MIT Sloan, most recently in spring of 2014. She collaborates actively with professors from a wide variety of universities, currently focusing on North America, Finland, Japan, and the UK. One of the key tools developed by OpLab is listed under the US Department of Commerce inventory for cost assessment . She has also been named as a professor of the Swiss Finance Institute because of her application of finance tools to the financial flows in supply chains.