Seminar Details

From sports to politics, from the stock market to prediction markets, from cancer detection to sequence completion, prediction is a big business. Proven track records of accurate predictions support claims of expertise and clairvoyance. Yet, what if forecasters could always cleverly cook their books to make their predictions look good in hindsight? What if we could always make predictions that are not just good, but in fact, unfalsifiable, simply by correcting prior mistakes? In this talk, I’ll give an overview of a line of work in online prediction, Defensive Forecasting, that provides simple strategies that achieve these guarantees.