基於高維度資料分解的空氣污染視覺化分析

Abstract

空氣污染是一個嚴重的全球環境問題，對人類健康和生態平衡造成嚴重影響。PM2.5是微粒物質的一個子集，直徑小於2.5微米，已經與嚴重的呼吸和心血管問題、土壤和水污染以及生態系統破壞相關聯。為了更好地了解PM2.5的來源和分佈，我們採用了一種類似PARAFAC的分解方法來分析台灣使用空氣盒子設備收集的空氣質量數據。這種方法允許識別導致某個地區和時間PM2.5濃度較高的因素，從而提供PM2.5分佈模式的洞察。為了增強對這些模式的分析，我們提出了一種通過可視化進行交互式多視圖分析的方法，以探索和理解複雜的數據集。這種方法旨在幫助更好地理解空氣質量，改進複雜數據集的分析和解釋，最終獲得更好的洞察和結果。

Air pollution is a serious global environmental issue that affects human health and ecological balance. PM2.5, a subset of particulate matter with a diameter of 2.5 micrometers or less, has been linked to severe respiratory and cardiovascular problems, soil and water pollution, and ecosystem disruption. To better understand the sources and distribution of PM2.5, we employed a PARAFAC-like decomposition method to analyze air quality data collected in Taiwan using airbox devices. This method allows for the identification of factors that contribute to high concentrations of PM2.5 in a given area and time, providing insights into the patterns of PM2.5 distribution. To enhance the analysis of these patterns, we propose an interactive multi-view analysis through visualization to explore and understand complex data sets. This approach aims to contribute to a better understanding of air quality and improve the analysis and interpretation of complex data sets, ultimately leading to better insights and outcomes.