Anonymized data supporting the paper "Transfer learning for mortality risk: A case study on the United Kingdom"

Dataset
Description

transfer_predictions_anonymized_UK
Minimal anonymized data set necessary to replicate the trasnfer methodology for the UK on age and gender grouping level. Variables present are age, gender, predicted mortality rates, deaths and exposure. The predictions are based on the synthetic dataset of UK - details can be found here (PDF) Transfer learning for mortality risk: A case study on the United Kingdom).

cmi_blended_rates
We utilize the ’16’ series 95mortality tables from Working Paper 154 for the evaluation purposes, given the absence of an actual UK portfolio for comparison. More details can be found on CMI website.