Estimating neighbourhood death rates using the random forest algorithm

olivier-collet-ZU62GJP-XC4-unsplash

Share:

Summary:

Recent decades have seen increasing evidence for inequality in mortality for different socio-economic groups in various national populations. In this article, published by the Institute and Faculty of Actuaries in their Longevity Bulletin, we construct a socio-economic index that models the ratio of actual vs. expected mortality in small neighbourhoods in England. The index is constructed using a Random Forest algorithm.

Publication:

Published in Institute and Faculty of Actuaries Longevity Bulletin issue 15. 

Click here to access the Institute’s Thought Leadership page with an article about the bulletin titled: The Machine Learning Issue.

Click here to access the bulletin directly, where the article is on page 22. 

Authors:

Andrew J.G. Cairns of The Maxwell Institute for Mathematical Sciences,  Department of Actuarial Mathematics and Statistics, School of Mathematical and Computer Sciences, Heriot-Watt University.

Jie Wen of Lloyds Banking Group.

Torsten Kleinow of Research Centre for Longevity Risk, Faculty of Economics and Business, University of Amsterdam.

 

Related research projects