Synthetic Data and the Insurance Industry: Impacts and Opportunities

Register

Part of the Synthetic Data Workshop Series

The insurance sector is a heavily regulated industry meaning that protecting customer data is of utmost importance to companies. However, this can pose a challenge when working on cutting-edge innovative data science solutions with often messy, imbalanced, and skewed tabular datasets. Synthetic data has the potential to shape the future of data science within the insurance sector; impacting the way models are built and tested to the way data is shared across teams, externally, and used for research or training purposes. As the Allianz Personal data science team, we have been investigating these potential uses of synthetic data through a series of internal research projects and guidelines. This talk will cover the synthetic data journey so far, talking about the challenges the team has faced, along with its current outlook and future hopes.

Agenda

• Who are we & what do we do?
• Why are we interested in synthetic data?
  • Data sharing (within company and externally within other group companies), modelling performance improvement, testing, training
• What work have we done internally with synthetic data?
  • Guidelines, imbalanced data, testing data
• Current outlook & future goals?
  • Privacy, performance improvement

Speakers (2)

Lorna Sumption

Lead Data Scientist LV = General Insurance

Dr Lorna Sumption obtained her PhD in developmental genetics before transitioning into the data science industry, joining Allianz Personal/LV=GI in 2020. As a lead data scientist, she has worked on several research-led projects including all phases of synthetic data work carried out within the team. These phases included investigating the practicalities of using synthetic data to address imbalanced datasets and for testing purposes.

Merve Alanyali

Head of Data Science Academic Partnerships and ResearchLV = General Insurence

Register