Statistical Modelling of Infectious Disease and Environmental Systems

传染病传播的模型对于维护人类、动物和植物的健康，以及维护经济安全和国际贸易都很重要。了解哪些因素影响疾病的传播，或者哪些人更有可能被感染，对于预防和控制疾病爆发至关重要。一类被称为“个体水平模型”的统计模型通过考虑人口中每个个体的信息来描述传染病在人口中的传播。这些个体可以是个人、植物、房屋或农场，每个人都有自己独特的信息集（如地理位置、是否接种过疫苗等）。个体水平模型使用这种非常具体的个体水平信息来描述传染病如何在空间和时间上传播。这些模型的优点是它们非常灵活，并且允许个体是异质的（例如，不完全相同）的群体。然而，这些模型是有限的，因为它们需要大量的计算时间，这在疾病爆发时可能是不可接受的，因为需要快速了解疾病是如何发生的。这里描述的研究计划旨在开发可以减少这种计算负担的方法，同时仍然准确地识别可能影响疾病传播的因素。此外，该研究计划旨在开发一个个体水平模型的子组，可以纳入缺失或可能不准确的个体水平数据，这是公共卫生和其他此类数据中经常出现的现象。这项研究计划的另一个目标是与环境影响监测倡议有关（即，阿尔伯塔的联合联邦和省政府油砂监测计划）。就像监测疾病爆发的想法一样，为了确定人类活动是否对野生种群产生了环境影响（即，鱼类），有必要识别这些种群何时发生显著变化（即，大小、体重、肿瘤发生率、繁殖成功率等）。用于监测环境影响的监测技术需要能够适应环境的自然变化，并区分由于工业/人类影响而观察到的影响，以及不属于生态系统中始终存在的自然变化的影响。该研究计划旨在开发用于监测环境影响的统计模型，这些模型可以识别显著的人口变化，同时结合有关周围环境和可能与这些变化有关的人类活动的信息。这一信息对于发现工业活动造成的影响和评估任何此类影响的程度至关重要。

Modelling the spread of infectious diseases is important for maintaining human, animal, and plant health, as well as maintaining economic security and international trade. Understanding what factors affect the spread of a disease, or which individuals are more likely to become infected is critical in preventing and controlling a disease outbreak.***A class of statistical models, known as “individual-level models”, describe the spread of an infectious disease through a population by considering information for each individual within that population. These individuals could be individual people, plants, houses, or farms, each with their own unique set of information (such as geographical location, whether they have been vaccinated, etc.). Individual-level models use this very specific, individual-level information to describe how an infectious disease may spread across both space and time. The advantage these models is that they are very flexible, and allow for a population in which individuals are heterogeneous (e.g. not all the same). However, these models are limited in that they require a large amount of computational time, which may be unacceptable in the event of a disease outbreak when it is necessary to quickly understand how a disease.******The research program described here aims to develop ways in which this computational burden can be reduced, while still accurately identifying factors which may influence the spread of the disease. Furthermore, this research program intends to develop a subgroup of individual-level models that can incorporate missing or potentially inaccurate individual-level data, a phenomenon that often arises in public health and other such data.******Another objective of this research program is related to environmental effects monitoring initiatives (i.e., the joint federal and provincial government oil sands monitoring program in Alberta). Like the idea of surveillance for a disease outbreak, to determine whether human activity is having an environmental impact on wild populations (i.e., fish), it is necessary to identify when there are significant changes in these populations (i.e., size, weight, tumour incidence, reproductive success, etc.). The surveillance techniques for monitoring environmental effects need to be able to adapt to natural changes in the environment, as well as distinguish observed effects that are due to industrial/human influence and are not within the natural variation that is always present within an ecosystem. This proposed research program aims to develop statistical models for monitoring environmental effects that can identify significant population changes, while incorporating information about the surrounding environment and human activities that may be related to these changes. This information is pivotal in detecting impacts caused by industrial activities, and assessing the extent of any such impacts.