25 June 2021
Congress Centre Bernardin, Portorož, Slovenia
UTC timezone
Internal satellite conference of the 8th European Congress of Mathematics - 8ECM

8 ECM / 8 ECM minisymposia

Below we have listed some of the minisymposia organized within 8 ECM that might also be interesting for insurance companies employees.

 

Mathematical challenges in insurance (MS - ID 41)

Statistics towards Industrial Mathematics (MS - ID 12)

Probabilistic approaches for studying blockchain dynamics (MS - ID 24)

 

The detailed description  is given below.

 

------------------------------------------------------------------------------------------

Mathematical challenges in insurance (MS - ID 41)

Organizer

Mihael Perman (University of Primorska, Slovenia), mihael.perman@fmf.uni-lj.si

Co-Organizers

Bor Harej (PRS Prime Re Solutions AG), bor.harej@prs-zug.com
Tomaž Košir ( University of Ljubljana), tomaz.kosir@fmf.uni-lj.si
Claudia Klüppelberg (TU München), cklu@tum.de
Thomas Mikosch (University of Copenhagen), mikosch@math.ku.dk
Ermanno Pitacco ( University of Trieste), ermanno.pitacco@deams.units.it

Description

Risk modeling is an essential tool for the insurance industry.  Recent changes in regulation on European level, the global reach of the European insurance industry and changing financial markets  along with strong  competition have created new challenges for the industry and for mathematicians.  A liberal regulatory regime allows insurance companies to develop compehensive risk management systems through internal models.  Such models are complex and involve advanced mathematical methods. The integration of statistical data processing, development of actuarial models and management decisions is a challenging problem. 

The availability of big data and advanced methods of machine learning is a new frontier for the insurance industry.  Applications range from detecting fraudulent practices, segmentation of customers, advanced pricing models to assisted financial decisions.  Fundamental mathematical questions of reliability and replicability of such applications must be addressed not only in insurance but in all applications of machine learning. 

The speakers will discuss modeling challenges and more fundamental questions of machine learning applications.  There has been a rising trend in weather related claims stretching the capacity of (re)insurance to its limits.  Complex mathematical models to adjust for trends and assist with pricing in non-life insurance is yet another challenge for mathematical modelling.  Some new advances in this respect will be presented.

-----------------------------------------------------------------------------------------------------------------

Statistics towards Industrial Mathematics (MS - ID 12)

Organizer

John Shawe-Taylor (University College London, Knowledge 4 All Foundation, UK), jst@cs.ucl.ac.uk

Co-Organizers

Benjamin Guedj (University College London, UK), benjamin.guedj@inria.fr
Marc Deisenroth (University College London, UK), m.deisenroth@ucl.ac.uk
Maria Florina Balcan (Carnegie Mellon University, USA), ninamf@cs.cmu.edu
Avrim Blum (TTIC, USA), avrim@ttic.edu

Description

 

Artificial Intelligence aims to understand and reproduce intelligent behaviour and decision making in computing systems. There has been a burgeoning interest in the topic following high profile successes such as the AlphaGo system that was able to defeat leading human players in the game of Go, widely regarded as a yardstick of human intelligence. This development did not come from nowhere, but rather built on developments in machine learning, reinforcement learning, statistical learning, in many cases supported by various mathematical and statistical foundations.

This minisymposium aims to bring together mathematical work that explores the underpinnings of all aspects of artificial intelligence, including (but not limited to) statistical learning theory, empirical process theory, stochastic control, Bayesian inference, decision theory, data compression, information theory, natural language processing, deep learning, model reduction, combinatorial search, network analysis, etc.

---------------------------------------------------------------------------------------------------------------------

Probabilistic approaches for studying blockchain dynamics (MS - ID 24)

Organizer

Frédérique Robin (INRIA Rennes-Bretagne Atlantique, France ), frederique.robin@inria.fr

Co-Organizers

Emmanuelle Anceaume (IRISA), emmanuelle.anceaume@irisa.fr
Bruno Sericola (INRIA), bruno.sericola@inria.fr

Description

 

Blockchain is an innovative technology that has the potential to transform many areas such as finance, logistics, and connected object management. It provides a means for exchanging or storing information secured by cryptography, without the need of a centralized trusted authority. 

The lack of a global trusted third party is bypassed by relying on a publicly readable and writeable blockchain, a data structure in which all transactions are gradually added through the creation of cryptographically linked blocks.

Initially introduced in 2008, the Bitcoin cryptocurrency system offers to its users the possibility to transfer funds (in Bitcoins) securely without a banking authority. Since then, new blockchain models have been introduced: the blockchain technology is still in the development phase and an open field of research. To ensure the viability of candidate improvement solutions, mathematical models and analyses tackling different blockchain aspects (consensus, forks, block creation or block spreading time, etc.), have been proposed during the last decades. 

This mini-symposium aims at bringing together researchers active on blockchain analysis, and allowing the interaction between different expertises: Markov chain, game theory, fluid limits as well as on different types of stochastic processes. It also aims at gather mathematicians analyzing blockchains through a theoretical approach to those applying numerical techniques, and thus it will encourage new collaborations.