Bootstrapping (Reinforcement Learning)
bootstrapping in RL means that you update a value based on some estimates and not on some exact values.
The bootstrapping is done via the Bellman Update.
“TD learning methods update targets with regard to existing estimates rather than exclusively relying on actual rewards and complete returns as in MC methods. This approach is known as bootstrapping.” Source
Bootstrapping: the update involves an estimate
- DP Bootstraps
- MC does not bootstrap
- TD Boostraps
Bootstrapping main ideas: https://www.youtube.com/watch?v=Xz0x-8-cgaQ (this is different though, its statistics)