Big data problems on discovering and analyzing causal relationships in epidemiological data

This research focuses on learning causal relationships on epidemiological data. We introduce the research need for causal reasoning and address one of the big data problems in epidemiology by showing the complexity of causal discovery and analysis in an observational epidemiological dataset. We also provide several computational methods of solving the problems including building a framework of causal reasoning on epidemio-logical dataset, improved algorithms for local causal discoveries, and the conceptual design of subgraph decompositions. This research further discusses how these approaches we have made are related to epidemiology. Through this research, we are able to more efficiently and effectively discover and analyze causal relationships in a big dataset of epidemiology.