Lund Albert M, Gouripeddi Ramkiran, Facelli Julio C
Online J Public Health Inform. 2020 Jul 30;12(1):e9. doi: 10.5210/ojphi.v12i1.10588. eCollection 2020.
Human activity encompasses a series of complex spatiotemporal processes that are difficult to model but represent an essential component of human exposure assessment. A significant empirical data source, like the American Time Use Survey (ATUS), can be leveraged to model human activity. However, tractable models require a better stratification of activity data to inform about different, but classifiable groups of individuals, that exhibit similar activity sequences and mobility patterns. Using machine learning algorithms, we developed an unsupervised classification and sequence generation method that is capable of generating coherent and stochastic sequences of activity from the ATUS data. This classification, when combined with any spatiotemporal exposure profile, allows the development of stochastic models of exposure patterns and records for groups of individuals exhibiting similar activity behaviors.
人类活动包含一系列复杂的时空过程，这些过程难以建模，但却是人类暴露评估的重要组成部分。像美国时间使用调查（ATUS）这样重要的经验数据来源可用于对人类活动进行建模。然而，便于处理的模型需要对活动数据进行更好的分层，以便了解不同但可分类的个体群体，这些个体群体表现出相似的活动序列和移动模式。利用机器学习算法，我们开发了一种无监督分类和序列生成方法，该方法能够从ATUS数据中生成连贯的和随机的活动序列。这种分类与任何时空暴露概况相结合，能够为表现出相似活动行为的个体群体开发暴露模式和记录的随机模型。
