Timeline for Reinforcement learning: How to deal with illegal actions? [duplicate]
Current License: CC BY-SA 4.0
10 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Sep 27, 2020 at 22:41 | history | closed | nbro | Duplicate of How should I handle invalid actions (when using REINFORCE)? | |
| Jun 14, 2020 at 12:21 | review | Close votes | |||
| Jun 19, 2020 at 3:04 | |||||
| Aug 23, 2019 at 3:05 | review | Close votes | |||
| Aug 23, 2019 at 20:48 | |||||
| Aug 14, 2019 at 20:51 | answer | added | Terry T. | timeline score: 1 | |
| Aug 14, 2019 at 14:40 | review | Close votes | |||
| Aug 14, 2019 at 19:40 | |||||
| Aug 14, 2019 at 13:23 | answer | added | mglss | timeline score: 1 | |
| Aug 14, 2019 at 13:17 | comment | added | agold | I would ignore forbidding actions (e.g. act like 0) in your environment (simulator), since - for as far I understood - they cause the system to be in an invalid state. Nevertheless, you probably also should put a (small) negative reward for those actions. | |
| Aug 14, 2019 at 11:34 | comment | added | Brale | ignoring forbidden actions is the right approach, only consider the allowed actions at specific timestep. | |
| Aug 14, 2019 at 11:25 | review | First posts | |||
| Aug 17, 2019 at 19:35 | |||||
| Aug 14, 2019 at 11:20 | history | asked | JH Lee | CC BY-SA 4.0 |