This is my implementation of the Q-Learning algorithm. It is in the category of Machine Learning (ML). Or even more specifically Reinforced Learning (RL). Where the "agent" learns about its enviroment by probing different parts (states) of it.
-
In its current state is able to:
- [v] Work in user define enviroment with no restriction on actions definition.
- [v] Finding optimal path to a given goal from every point in the enviroment.
- [v] Supports multi-objective learning.
- [!] For each objective the enviroment must be redefined.
- [SOON] Saving already trained models to files, and loading them later.
For more info check the Documentation, or one of the links bellow.
If you are intrested: Wikipedia: https://en.wikipedia.org/wiki/Q-learning
Not-So-Complex explanation: https://mnemstudio.org/path-finding-q-learning-tutorial.htm
YouTube[Siraj Raval]: https://www.youtube.com/watch?v=aCEvtRtNO-M
Siraj Raval's YouTube channel is dedicated to computer/math science and has many lecture like videos.
PS: Hope I'm not doing 'github' too bad.