Keynote | Technical
Friday 17th | 12:00 - 12:30 | Theatre 25
Where statistical errors come from, how they cause us to make bad decisions, and what to do about it.
Keywords defining the session:
- decision making
Takeaway points of the session:
- Learn to anticipate errors in data, models, and predictions we make
- Learn how to estimate uncertainty in a wide range of data science applications: surveys, product tests, and forecasts.
With the surging use of data for decision making, we are forced to confront that there is error in almost everything we measure and every prediction we make. While most people learn about Type I/II errors and are comfortable evaluating machine learning models performance, there are many more ways for errors to propagate in data science systems that cause bad decisions or product experiences. In this talk, I provide a generic framework for thinking about errors in data science systems. I show that anticipating errors and quantifying our uncertainty can help us use data more effectively.