Scholarlink

Environmental risk assessment is vital for understanding hazards to public health and ecosystems, with air pollution presenting significant challenges. Despite advances in urban monitoring, pollutants are often studied in isolation, neglecting their complex interdependencies. We develop a non-homogeneous Vector Auto Regressive - Hidden Semi-Markov Model that incorporates external covariates influencing state dynamics. We adapt financial risk measures to the environmental domain and refine Shapley-value methods for fair risk attribution. This multivariate framework enhances interpretability and deepens insight into environmental hazards.

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

Environmental Risk Assessment for Multivariate Time-Series of Air Polluttants: A Shapley Value Approach to Identifying Key Contributors

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Pierfrancesco Alaimo Di Loro,
  • Francesco Lagona,
  • Antonello Maruotti

摘要

Environmental risk assessment is vital for understanding hazards to public health and ecosystems, with air pollution presenting significant challenges. Despite advances in urban monitoring, pollutants are often studied in isolation, neglecting their complex interdependencies. We develop a non-homogeneous Vector Auto Regressive - Hidden Semi-Markov Model that incorporates external covariates influencing state dynamics. We adapt financial risk measures to the environmental domain and refine Shapley-value methods for fair risk attribution. This multivariate framework enhances interpretability and deepens insight into environmental hazards.