Online Program

Are Data Science and Analytics Just New Names for Statistics? (302946)

Keywords: Data science, analytics, predictive models

In the first part of this presentation, we will provide an overview of data science and analytics and show how they relate to traditional statistics, machine learning, and data mining. One feature of this new setting is the ubiquity of large data sets, leading to the current buzz word—Big Data. The large numbers of both observations and variables lead to the need for more flexible and automated ways for modeling and prediction. This, in turn, increases applicability of predictive algorithms of machine learning as improvements over the traditional statistical models. In the second part of this presentation, we will use a simple data set to illustrate the gamut of models and algorithms, from the simple to the more complex, including bagging, boosting, and random forests. This is a nontechnical presentation, with the concepts shown mostly through graphs. The goal is to help practicing statisticians see how their experience and expertise fits into this new landscape.