Gradient boosted multi-population mortality modelling with high-frequency data

2025-07-14

AI Summary

High-frequency mortality data exhibit complex temporal structures, rendering conventional models inadequate for accurately capturing short-term trends and seasonality. Method: This paper proposes a novel mortality modeling framework that integrates gradient boosting with multi-population stochastic mortality modeling. Specifically, the Li-Lee model is innovatively incorporated as a weak learner within a gradient boosting ensemble; population clustering is optimized based on mortality improvement rates and seasonal intensity to refine subgroup definitions. The method operates directly on weekly mortality data and is systematically evaluated across a 30-country sample. Contribution/Results: The proposed model consistently outperforms mainstream benchmarks in goodness-of-fit, short-term forecasting accuracy, and robustness across diverse cluster configurations. It establishes a scalable, interpretable, machine learning–enhanced paradigm for high-frequency mortality modeling, bridging actuarial theory with modern ensemble learning techniques.

Technology Category

• Machine Learning: Ensemble Methods
• Reasoning under Uncertainty: Graphical Models
• Search and Optimization: Learning to Search

Application Category

• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

High-frequency mortality data remains an understudied yet critical research area. While its analysis can reveal short-term health impacts of climate extremes and enable more timely mortality forecasts, its complex temporal structure poses significant challenges to traditional mortality models. To leverage the power of high-frequency mortality data, this paper introduces a novel integration of gradient boosting techniques into traditional stochastic mortality models under a multi-population setting. Our key innovation lies in using the Li and Lee model as the weak learner within the gradient boosting framework, replacing conventional decision trees. Empirical studies are conducted using weekly mortality data from 30 countries (Human Mortality Database, 2015--2019). The proposed methodology not only enhances model fit by accurately capturing underlying mortality trends and seasonal patterns, but also achieves superior forecast accuracy, compared to the benchmark models. We also investigate a key challenge in multi-population mortality modelling: how to select appropriate sub-populations with sufficiently similar mortality experiences. A comprehensive clustering exercise is conducted based on mortality improvement rates and seasonal strength. The results demonstrate the robustness of our proposed model, yielding stable forecast accuracy under different clustering configurations.

Problem

• Modeling high-frequency mortality data with complex temporal structures
• Improving forecast accuracy using gradient boosting in multi-population settings
• Selecting sub-populations with similar mortality experiences for robust modeling

Innovation

• Integrates gradient boosting with mortality models
• Uses Li and Lee model as weak learner
• Clusters populations by mortality improvement rates