Scholarlink

<p>In this paper, a&#xa0;method for knowledge-based optimization of vaccination assignments is proposed, which combines multiobjective optimization algorithms with counter-epidemic strategies known from epidemiology. In the paper, a&#xa0;model based on real-life illness costs is used, which allows taking into account the age of individuals exposed to a&#xa0;simulated epidemic. Using this model, strategies based on the age and on the graph node degree (the number of contacts each individual has) are studied. The optimization algorithms work on a&#xa0;graph which represents a&#xa0;“known” network of contacts and the solutions are subsequently used for controlling an epidemic spreading on another graph representing an “actual” network of contacts. The optimized solutions are tested on “actual” graphs with a&#xa0;varying degree of overlap with the “known” graph on which the optimization algorithms work. In the experiments, the age-based strategy was found to perform best, and the degree-based strategy turned out to be the second-best approach. Both these strategies outperformed other approaches not using knowledge from epidemiology to improve optimization results.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Knowledge-based optimization in epidemics prevention

  • 期刊论文

  • 发表日期：

    2025年1月1日

查看全文
  • Krzysztof Michalak

摘要

In this paper, a method for knowledge-based optimization of vaccination assignments is proposed, which combines multiobjective optimization algorithms with counter-epidemic strategies known from epidemiology. In the paper, a model based on real-life illness costs is used, which allows taking into account the age of individuals exposed to a simulated epidemic. Using this model, strategies based on the age and on the graph node degree (the number of contacts each individual has) are studied. The optimization algorithms work on a graph which represents a “known” network of contacts and the solutions are subsequently used for controlling an epidemic spreading on another graph representing an “actual” network of contacts. The optimized solutions are tested on “actual” graphs with a varying degree of overlap with the “known” graph on which the optimization algorithms work. In the experiments, the age-based strategy was found to perform best, and the degree-based strategy turned out to be the second-best approach. Both these strategies outperformed other approaches not using knowledge from epidemiology to improve optimization results.