What is it about?
Knowledge-Driven model: KD model refers to a model derived from domain knowledge, also known a mechanistic model (Fan et al., 2015; Kang et al., 2018). Data-Driven model: DD model refers to a model that can be learn from data without using any domain knowledge (Fan et al., 2015; Kang et al., 2018).
Featured Image
Why is it important?
Advantage and disadvantage of KD and DD models. KD models: Pros. Contain physically interpretable parameters Show strong explanatory power Cons. Limited in predictive ability DD models: Pros. Avoid complex modeling of infectious disease mechanism Exhibit a high predictive ability Cons. Lack interpretability because of their black-box nature
Perspectives
Motivation: Most existing epidemic prediction models were established from a single perspective (i.e., KD or DD model) and the predictive performance is not good (Zheng et al., 2020). Inspired by knowledge-and-data-driven modelling (KDDM) approaches (Fan et al., 2015, 2018). To develop an epidemic prediction model with high accuracy and strong interpretability for making more effective prevention and control policies during pandemic crises.
Xing-Rong Fan
Chongqing Technology and Business University
Read the Original
This page is a summary of: Stacking based prediction of COVID-19 Pandemic by integrating infectious disease dynamics model and traditional machine learning, August 2022, ACM (Association for Computing Machinery),
DOI: 10.1145/3561801.3561805.
You can read the full text:
Contributors
The following have contributed to this page
