Timeline for What are other ways of handling invalid actions in scenarios where all rewards are either 0 (best reward) or negative?
Current License: CC BY-SA 3.0
9 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Aug 24, 2020 at 12:25 | comment | added | d56 | @BlueMoon93, I don't understand how changinng reward interval from [-10,0] to [0,+10] would improve learning? In the end, it is just shifted. In the first case the agent would be motivated not to do stupid actions in order to get closer to 0... | |
| Jun 17, 2020 at 9:57 | history | edited | CommunityBot |
Commonmark migration
|
|
| May 29, 2017 at 13:53 | vote | accept | AlexGuevara | ||
| May 29, 2017 at 12:47 | comment | added | BlueMoon93 | By setting an end reward, you just motivate the agent to finish the task (since he gets the best reward out of finishing it). You still need the negative rewards, of course, to make him avoid wasting resources. With penalties and an end-game reward, you get the best of both worlds. | |
| May 29, 2017 at 12:46 | comment | added | BlueMoon93 | You can use some heuristics to start with a network whose weights are not completely random. It is called transfer learning. But it's hard to do because it implies you knowing the values somehow (usually, through previous learning stages). You also cannot just set some specific values and expect others to remain unchanged. As you adjust weights to get some actions to -inf, other action's will get very strange reward values, which will most likely disturb training. Once you start optimizing the network, when you explore, backprop will just optimize the weights and set everything back. | |
| May 29, 2017 at 12:30 | comment | added | AlexGuevara | I do not quite understand why initially setting the network weights would defeat the backpropagation purpose? Why is it unfeasible to use quick, greedy heuristics to estimate the value of a certain state, can you elaborate? Setting an end reward motivates the agent to reach the end, but it would still need the negative rewards to keep it from using resources it does not need, right? There is no time limit: I do not care if it takes longer to complete a task, as long as it uses as few resources as possible. | |
| May 29, 2017 at 11:08 | history | edited | BlueMoon93 | CC BY-SA 3.0 |
added 944 characters in body
|
| May 29, 2017 at 11:03 | review | First posts | |||
| Jun 7, 2017 at 14:19 | |||||
| May 29, 2017 at 11:02 | history | answered | BlueMoon93 | CC BY-SA 3.0 |