Giornate di Studio sulla Popolazione (Popdays), Giornate di Studio sulla Popolazione 2019

Font Size: 
Mortality forecasting. Machine Learning Assisted Approach (MLAA)
Andrea Nigri

##manager.scheduler.building##: Velodromo - Bocconi University
##manager.scheduler.room##: N03
Date: 2019-01-24 04:30 PM – 06:00 PM
Last modified: 2018-12-26

Abstract


In the eld of mortality, the Lee-Carter (L-C) based approach is the best wayto forecast mortality rates. Since the rst version of the model in 1992, scholarshave developed dierent versions of it, so we could therefore dene an "L-Cmodel family" that includes all developments of it. Nevertheless, the rst for-mulation of 1992, remains the benchmark for comparing the performance offuture models. The main aim of this thesis is to try to ll a gap between de-mography methodology and statistical learning eld. Indeed using data fromHuman Mortality Database we will attempt to integrate the L-C model withmachine learning approach, in particular by using the Neural Network (NN).

Keywords


mortality, methods, machine learning