Abstract:
|
The term data science is often associated with machine learning, which is powerful in its ability to find patterns and relationships in data for purposes of prediction and classification. Indeed, there are those that use the terms interchangeably. However, in the past five years the field of data science has seen the value of designed experiments for purposes of advertising, product development, product improvement, user engagement, customer acquisition and customer retention to name just a few. The importance of experimentation to the field of data science is evident in the plethora of data scientist job ads that explicitly highlight the need for expertise in the design and analysis of experiments. This increased emphasis of experimentation in the field of data science represents an opportunity and a challenge for DOE educators. The types of classes we offer, and the manner in which our traditional classes are offered, need to adapt. In this talk I discuss the relevance of experimentation to the field of data science and my experience developing curriculum and teaching experimental design in this unconventional domain.
|