Timeline for Practical differences between the different RL architectures
Current License: CC BY-SA 4.0
3 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Sep 12, 2023 at 16:31 | answer | added | pi-tau | timeline score: 1 | |
| Sep 12, 2023 at 15:55 | comment | added | Alberto | those are algorithms, not architectures, and they differ in the ability to learn... however, if your reward signal is a +1/-1 after 382739857 actions, it's veeery hard for all of them to learn, as there is no "supervision" signal to optimize for... you might want to look at keywords like "reward shaping" and "intrinsic curiosity" | |
| Sep 12, 2023 at 15:20 | history | asked | Erik Storm | CC BY-SA 4.0 |