luccabz

14 ms·

1. 1.▲
   by luccabz 10mo ago

   we should: 1. train with a cutoff date at ~2006 2. simulate information flow (financial data, news, earnings, ...) day by day 3. measure if any model predicts the 2008 collapse, how confident they are in the prediction and how far in advanc
2. 2.▲
   A ~2000 Elo Python chess engine(github.com)

   3 points by luccabz 11mo ago|1 comments