Giters
dquail
/
PredictiveKnowledge
Learning predictive representations of knowledge
Geek Repo:
Geek Repo
Github PK Tool:
Github PK Tool
Stargazers:
Watchers:
3
Issues:
8
Forks:
1
dquail/PredictiveKnowledge Issues
It looks like GTD does not produce a RUPEE estimate, or a return error, or UDE.
Updated
6 years ago
Comments count
1
Cumulant over-ride for distance_to_touch is just a constant
Updated
6 years ago
Possible bug: overriding the learning functionality for the touch-adjacent learning
Updated
6 years ago
Refactor policy selection so that it uses a numerical representation, rather than a dict
Updated
6 years ago
Need ability to pickle and unpickle GVF weights
Closed
6 years ago
Comments count
3
Ability to control the agent from the user interface
Closed
6 years ago
Comments count
2
Add bias bit
Updated
6 years ago
Add game video output to display
Updated
6 years ago