3 ms·
Double-Loop Learning
- VivaTechnics 1y agoAwesome! Technically, we could say? (1) Single-loop: fixes actions within fixed rules, like Reinforcement Learning. (2) Double-loop: questions and adapts the rules, somewhat like Meta-Reinforcement Learning.
- 01HNNWZ0MV43FF 1y agoFive Whys is supposed to get you to this
- treetalker 1y agoSimilar/related, for anyone who hasn't heard of it yet: the OODA loop (https://en.m.wikipedia.org/wiki/OODA_loop https://en.m.wikipedia.org/wiki/OODA_loop)