Abstract:
|
Less than two years ago, I thought experimental design was important and interesting within its traditional domains, but largely irrelevant to analytics. I was wrong; I now believe that design has become more important, not less important, in this big data era. In this talk, I will provide an overview of the connections between statistically designed experiments and aspects of machine learning and data science, with a particular eye on establishing causal relationships by combining observational and experimental data.
|