Alphabetical list of class partners and collaborators:
- Athena Health
- Caesars Entertainment
- Deloitte
- Diffeo
- Ebay
- Harvard
- Hubway/MAPC
- IBM
- KBA/DARPA
- MIT
- Nationwide
- Risk Management Services
- Sense
- Siemens
- Starbucks
In this class, we will build a way of thinking about quantitative problems regardless of their seeming complexity. The class covers all stages of a data analysis problem: (1) setting up the problem, (2) designing a method to solve it, (3) implementing the related computation, (4) presenting the findings. We will uncover the components of each stage, demonstrate the interaction between all stages (1-4), and discuss their practical implementation.
In the lectures, we will apply this way of thinking to computing algorithms, traditional statistical models, latest advances in statistical research, and case studies from industry. We will discuss how efficient programming enables problem solving for data analysts. We will showcase the recent advances in visualization and their connection to the statistical world.
Problem sets guide students through practical computing problems, inviting to fill in missing parts of the code and highlighting the fundamental statistical questions along the way. There is also an interactive visualization component in each homework for extra credit.
Final projects are based on data-driven problems in research and industry. Working on the final project, the students will go through all 4 stages of a data analysis problem. Course partners view final projects as high priority and look forward to the solutions generated by the students.