Insurance Data Science conference 2024

24 Jun 2024

The conference took place at Stockholm University, Department of Mathematics, Campus Albano, House 1, Stockholm, 17 - 18 June 2024

Download conference abstract and programme booklet

Keynotes

Katrien Antonio (Professor, KU Leuven & University of Amsterdam)
Björn Dalemo (CEO, Länsförsäkringar Alliance)
Mario V. Wüthrich (Professor, ETH Zürich)

17 June 2024

08:00 - 09:00 Registration

09:00 - 09:15 Room Hörsal 1: Opening remarks

09:15 - 10:15 Room Hörsal 1: Keynote 1 (Chair: Mathias Lindholm)

Björn Dalemo (Länsförsäkringar Alliance): Trust and stand by algorithmic decisions - a management perspective

10:15 - 11:15 Regular Session 1

Room Hörsal 1: Generative AI in insurance 1 (Chair: Jürg Schelldorfer)

Dylan Liew (a) and Malgorzata Smietanka (a,b) ((a) Institute and Faculty of Actuaries Federated Learning Working Party, (b) UCL): Hush hush: Keeping neural network claims modelling private, secret, and distributed using federated learning
Yves-Cédric Bauwelinckx (KU Leuven): Generating individual claims using generative adversarial networks
Louis DOUGE (SwissRe): Gen AI: Developing and deploying a specialized underwriting AI assistant

Room Lärosal 5: Machine learning & climate modelling (Chair: Filip Lindskog)

Ronald Richman (Old Mutual Insure, University of the Witwatersrand ): Claims modelling with climate data
Claudio Giancaterino (Intesa San Paolo Vita): Boost Climate Risk Modelling with Large Language Models Data Augmentation
Efren Hernandez (Management Solutions): Quantification of climate risk in Insurance

11:15 - 11:40 Coffee break

11:40 - 12:40 Lightning Session 1

Room Hörsal 1: Stream 1 (Chair: Markus Gesmann)

Alan Muro (Swiss Re Corporate Solutions): Enabling business steering and decision-making through volatility modelling
Finn-Erik Langeggen (Advisense): Monitoring and modelling ESG risks in insurance
Jakob Dambon (Swiss Re): Advancements in embedded insurance pricing models for cargo insurance
Despoina Makariou (University of St Gallen): A causal machine learning approach for estimating heterogeneous treatment effects in the primary catastrophe bond market
Daniel Knös (Guy Carpenter): Cloudburst hazard analysis – a GIS approach

Room Lärosal 5: Stream 2 (Chair: Rui Zhu)

Jimmy Hollén (The Swedish Financial Supervisory Authority): Machine learning and supervision
Erik Gustafsson (Advisense): Challenges and opportunities when implementing AI/ML in pricing and reserving
Eivind Borg (Analytika): Leveraging AI for DORA compliance assessment in insurance
Javier Calvo (Management Solutions): The impact of artificial intelligence on financial institutions
Sven Haadem (Aeda): Enabling advanced analytic capabilities by multiplying internal insurance data
Emil Bustos (Research Institute of Industrial Economics): How well do firms recover from idiosyncratic shocks? Evidence from insurance claims

12:40 - 13:40 Lunch

13:40 - 14:40 Regular Session 2

Room Hörsal 1: Machine learning & fairness (Chair: Mathias Lindholm)

Arthur Charpentier (UQAM, Montréal): Obtaining fair insurance premiums with multiple sensitive attributes
François Hu (Milliman Paris): Ensuring algorithmic fairness in insurance pricing: A multi-class problem perspective
Tessa Steensgaard (University of Copenhagen): Fair learning and testing for unfairness given protected features

Room Lärosal 5: Machine learning & reserving 1 (Chair: Martin Bladt)

Grainne McGuire (Optum): Measuring loss reserving uncertainty with machine learning models
Emil Hofman (University of Copenhagen / Alm Brand Group): A machine learning approach based on survival analysis for IBNR frequencies in non-life reserving
Oliver Lunding Sandqvist (University of Copenhagen & PFA Pension): Estimation subject to reporting delays and incomplete event adjudication with applications to disability insurance

14:40 - 15:40 Panel discussion (Room Hörsal 1)

Generative AI and new technologies in insurance (Chair: Ronald Richman)

Panel

Amélie Breitburd (CNP Assurances)
Louis Douge (Swiss Re)
Grainne McGuire (Optum)
Per Jakobsson (Finansinspektionen)

15:40 - 16:00 Coffee break

16:00 - 17:00 Regular Session 3

Room Hörsal 1: Machine learning & pricing 1 (Chair: Mario V. Wüthrich)

Łukasz Delong (University of Warsaw): Isotonic regression for variance estimation and its role in mean estimation and model validation
Anne van der Scheer (Perunum Actuarieel Advies): Credibility in network regression with sigma-hot encoding and weight balancing
Henning Zakrisson (Stockholm University): A tree-based varying coefficient model

Room Lärosal 5: Machine learning & life insurance (Chair: Salvatore Scognamiglio)

Luca De Mori (Bayes Business School): Mortality forecasting via multi-task neural networks
Gayani Thalagoda (University of New South Wales): Variable annuity portfolio valuation with Shapley additive explanations
Francesco Ungolo (University of New South Wales): A Dirichlet process mixture regression model for the analysis of competing risk events

17:00 - 18:00 Keynote 2 (Room Hörsal 1) (Chair: Arthur Charpentier)

Katrien Antonio (KU Leuven & University of Amsterdam): Insights from fine-grained mortality data

19:00 Conference dinner

18 June 2024

09:00 - 10:00 Room Hörsal 1: Keynote 3 (Chair: Filip Lindskog)

Mario V. Wüthrich (ETH Zürich): Experience rating

10:00 - 11:00 Regular Session 4

Room Hörsal 1: Generative AI and underwriting (Chair: Erik Gustafsson)

Benedikt Herwerth (Swiss Re): LLM-Powered information extraction for claim documents
Jayant Apte (Scor, Charlotte): Accelerated underwriting and underwriting with partial information
Per Wilhelmsson (Länsförsäkringar): Unsupervised learning for efficient underwriting

Room Lärosal 5: Machine learning & financial aspects of insurance (Chair: Boualem Djehiche)

Patrick Cheridito (ETH Zürich): Computing capital requirements with guarantees
Han Li (The University of Melbourne): Augmenting hierarchical time series through clustering: Is there an optimal way for forecasting?
Salvatore Scognamiglio (University of Naples Parthenope): Multiple yield curve modeling and forecasting using deep learning

11:00 - 11:30 Coffe break

11:30 - 12:30 Regular Session 5

Room Hörsal 1: Machine learning & pricing 2 (Chair: Can Baysal)

Paul Wilsens (KU Leuven): Reducing the dimensionality and granularity in hierarchical categorical variables
Jan Küthe (Akur8): Penalized regression - Between Credibility and GBMs
Mathias Lindholm (Stockholm university): On duration effects in non-life insurance pricing

Room Lärosal 5: Machine learning & reserving 2 (Chair: Grainne McGuire)

Gabriele Pittarello (University of Torino): Individual claims reserving using the Aalen-Johansen estimator
Martin Bladt (University of Copenhagen): Bootstrapped multi-states and model uncertainty
Selim Gatti (ETH Zürich): Modeling lower-truncated and right-censored insurance claims with an extension of the MBBEFD class

12:30 - 13:25 Lunch

13:25 - 14:25 Regular Session 6

Room Hörsal 1: Machine learning in insurance 1 (Chair: Katrien Antonio)

Agathe Fernandes Machado (UQAM, Montréal): Challenging calibration of insurance scoring algorithms
Charlotte Jamotton (Université catholique de Louvain): Variational autoencoder for synthetic insurance data
Can Baysal (Munich RE): Demand & elasticity modelling for P&C insurance pricing under various conditions

Room Lärosal 5: Machine learning in insurance 2 (Chair: Boualem Djehiche)

Meryem Yankol-Schalck (IPAG Business School): Auto insurance fraud detection: Leveraging cost sensitive and insensitive algorithms for comprehensive analysis
Giovanni Rabitti (Heriot-Watt University): Measuring unexplained variation in insurance data: a non-parametric approach based on global sensitivity indices
Mick Cooney (Describe Data): Using random portfolios for managing and assessing insurance risks

14:25 - 14:55 Lightning Session 2

Room Hörsal 1: Stream 1 (Chair: Mathias Lindholm)

Yousra Cherkaoui (CREST-ENSAE): Cyber risk modeling using a two-phase Hawkes process with external excitation
Emilio S. Guillén (Bayes Business School, City, University of London): Generalized additive models and functional gradient boosting with geometrically designed (GeD) splines: Application to insurance data
Antoine Burg (Université Paris Dauphine (CEREMADE) & SCOR): Leverage closed-form MLE for multivariate regression models: GLM-trees and actuarial applications

Room Lärosal 5: Stream 2 (Chair: Markus Gesmann)

Giulia Pucci (KTH Royal Instituto of Technology): Network-based optimal control of pollution growth
Adam Bedwell-Smith (Ki Insurance): Genetic algorithms for property binder portfolio optimisation
Soroush Amirhashchi (Plannet Insurtech Hub): Modeling mortality rates: Nonparametric Bayesian inference with gaussian Cox processes

14:55 - 15:10 Room Hörsal 1: Closing comments

Advisense is a leading governance, risk and compliance powerhouse, offering best-in-class services and tech solutions to the European industry.

Silver Sponsors

Mirai Solutions: Smarter analytics - better decisions

Swedbank: Swedbank Försäkring

Swiss Re: We make the world more resilient

Stockholm .Stockholm