Causal learning from high-dimensional observational data
Preetam Nandy · Repository for Publications and Research Data (ETH Zurich) · 2016
Many scientific questions are causal in nature.Observational studies are sometimes the only source of information for answering such questions, especially when intervention experiments are impractical or unethical.Moreover, even if intervention experiments are possible, readily available observational data can be used, for example, in prioritizing much more expensive and time-consuming experiments.