Federated learning is a new class of machine learning models pioneered by the likes of Google and Apple, which allows models to be trained in the cloud, but without requiring customer or user data to be sent or transmitted, in order to preserve their privacy, which grows ever more important in the age of GDPR, etc. Unlike the traditional machine learning modelling approach that collects raw data to train on a centralised model (‘taking the data to the algorithm’), federated learning keeps data stored locally and instead, sends the model to the users (‘taking the algorithm to the data’). Coupled with an added layer of unique encryption, this technique means data can be stored and kept privately, but still allow models to be built upon them.


In this session, the IFoA Federated Learning Working Party will demonstrate how this might apply to modelling insurance claims. Specifically how it might be possible for several different insurance companies to collectively build a combined, shared, claims frequency neural network (similar to the CMI for mortality), whilst keeping their data and predictions completely hidden from each other.


Speakers: 


Malgorzata Smietanka, UCL

Dylan Liew, Data Science Actuary, BUPA

Background on pandemics: important historic pandemics (the Black Death or the Plague of Justinian) have been (arguably) worse than COVID. Pandemics today: Engineered v natural; Deliberate v accidental


How can we learn from the burgeoning field of “existential risks”? Which pandemics are most likely to lead to human extinction? How does this help model pandemic risk on an insurance balance sheet?


What can we as actuaries do?


Speaker: Sanjay Joshi, Hymans Robertson

The Data Visualisation Working Party and the Supervised Learning Working Party are teaming up to present together their work on the French Motor Third Party claims dataset.


The talk will include:

  • Background on the French Motor Third Party dataset
  • Visualisation of the dataset using R, Python, and Tableau
  • The approach and results of the baseline ZIP model
  • The approach and results of GBMs, Tree models, Neural Networks, and alternative models – how these models work in-depth and/or how we trained them
  • Overall results and discuss the potential next steps

Speakers: 

Eilish Bouse, Grant Thornton

Neptune Jin, BGL Insurance Group

Jonathan Bowden,  Hyperexponential

Afzaal Ahmed

  • How do you leverage your unique talents to stand out from the crowd?
  • How do you get comfortable stepping up when you don’t fit the “accepted mould” or usual stereotype?
  • How do you overcome hurdles to building a senior executive and/or NED career?
  • How can you be an ally to others on their journey?

In this interactive session, our three expert panelists will assist you with top tips on how to market your unique perspective to successfully become a senior executive or NED. With expertise in the Insurance & Financial Services sector, as well as Diversity, Equity and Inclusion, they will cover key topics such as networking, neurodiversity, executive search, career management, and allyship. 


You’ll leave with the tools to overcome career hurdles and achieve your ambitions! 


Speakers: 

Barbara Schonhofer, MBE, Executive Chair Global of ISC Group

Fiona Hathorn, CEO of Women on Boards UK

Mark Freed,  CEO E2W

March 2020 was the last in-person professional skills session we presented.  Since then we have seen major changes to where and how actuaries work and consequently changes in how we interact with our customers, clients and colleagues.

There have been new challenges of isolation, communication, company loyalty, conflicts and many others. Many of these changes will continue to some extent going forward as many of us adapt to a hybrid working environment. Throughout all this it is imperative we remain professional in all we do.

What additional aspects do we need to consider to achieve this?  How can the IFoA’s Regulatory approach support members in their endeavours to provide the best service to the users of actuarial work?

This will be a lively interactive session led by Malcolm Slee, Chair of the IFoA’s Professional Skills Sub-committee and an IFoA Regulatory Solicitor, where we consider these issues with the aid of a video case study and other material. 


Speakers:

Malcolm Slee, Chair of the IFoA’s Professional Skills Sub-committee

Emma Gilpin, IFoA

The Data and Modelling workstream of the Institute and Faculty of Actuaries Mental Health Working Party is publishing a paper in Q2 focused on data analysis and underwriting of mental health conditions in life and protection insurance products. This talk is focused on exploring potential avenues for improvements in data collection, changes to how mental health risks are modelled and considerations of innovative product designs.


Speakers:


Lisa Balboa, Hannover Re,

Joseph Wilson, RGA,

Serena Soong, Legal & General

Maryse Nashime, Partner Re

Cillian Tierney, Chief Medical Underwriter UKI/I at Partner Re

Day 1 of the IFoA Conference 2022

March 2020 was the last in-person professional skills session we presented.  Since then we have seen major changes to where and how actuaries work and consequently changes in how we interact with our customers, clients and colleagues.


There have been new challenges of isolation, communication, company loyalty, conflicts and many others.


Many of these changes will continue to some extent going forward as many of us adapt to a hybrid working environment.


Throughout all this it is imperative we remain professional in all we do.


What additional aspects do we need to consider to achieve this?  How can the IFoA’s Regulatory approach support members in their endeavours to provide the best service to the users of actuarial work?


This will be a lively interactive session led by Malcolm Slee, Chair of the IFoA’s Professional Skills Sub-committee and an IFoA Regulatory Solicitor, where we consider these issues with the aid of a video case study and other material. 


Speakers:


Malcolm Slee, Chair of the IFoA’s Professional Skills Sub-committee

Emma Gilpin, IFoA