An evidential-reasoning based model for probabilistic inference with uncertain data acquired from different data sources

摘要

This research aims to develop a new model for identifying asthma control steps in the framework of the Evidential Reasoning (ER) rule and to address the uncertainty issue related to prior distributions shown in datasets routinely generated from medical practices. The ER rule is applied to combine multiple pieces of evidence in a recursive fashion, with each piece of evidence acquired from an observable variable and represented as a probability distribution on hypothesis space. The proposed model has desirable flexibility in dealing with multiple pieces of evidence acquired from different data sources where the prior distributions of asthma control steps can be different.