Description: During the first week, faculty will present motivating data sets and research problems to the program participants. A crash course on R will be given as we will use R extensively. Specifically, we will start with data manipulation and data summary (in particular data visualization) so that participants can start playing with data from Day 1.
In this advanced data analysis program, we will introduce two powerful statistical techniques, namely, quantile regression and extreme value analysis for studying the "tails" of the distribution a random phenomenon (e.g., S&P 500 Index next quarter or daily maximum temperature in Clemson next 10 years). Questions such as: how to estimate the so-called Value-at-risk (VaR) of a financial time series or how to statistically infer the magnitude of a 100-year storm from a climate time series can be addressed by applying quantile regression and/or extreme value analysis properly.
Each participant can choose to work on an individual project or a group project. In the first few weeks, faculty will cover the needed materials for conducting their projects. Every week, the groups/individuals will update the entire group on what they learned, and what issues and difficulties they faced. Such updates will help keep everyone on task and focused.
Each participant will give a thirty-minute presentation (beamer encouraged but not required) on their project. This will help participants prepare for speaking on their projects during a job interview or presenting their research at conferences.
Prerequisites: Statistical Methods and Statistical Computing (required), and Advanced Statistical Methods (preferred)
Project faculty: