I am aware of keras, 'block' and a few others Python libraries for neural networks which do Reinforcement Learning among others. But is there a library than can make the task of visualisations easy? Particularly in 3D modelling of agents/environment,Seeing the simulations etc... I can see a few reinforcement learning videos online that show the simulated agent/environment but either they have made visual models from the ground up or used some other language/technology...(or they are very old). Open AI gym was one such platform where you can place and train your agents but I believe it's more for creating the learning algorithm than for visual simulation and environment.
1
-
Anurag, please edit your question, it is very hard to understand. Use proper interpunction and language (Not n where you mean and. What is nn?). Add links and references (e.g. those videos). Also: (what) are you willing to pay? Frankly, I don't even understand what visualizations you mean that are related to reinforcement learning.– user416Jun 15, 2016 at 12:37
Add a comment
|
3 Answers
You could also take a look at PyBrain, which implements some of the standard benchmark tasks and it allows for the experiments 3D visualization.
Here you can find some examples: https://www.youtube.com/watch?v=fEM7YDNonSE
See this link for testimonials about Psychopy, including one about the ease with which links to visualisation software can be made.
You could look at RL Toolkit, which gives examples of visualizations for mountain car and grid worlds.
http://incompleteideas.net/rlai.cs.ualberta.ca/RLAI/RLtoolkit/RLtoolkit1.0.html
-
I posted the same question on SO. got openAI gym as a recommended library. Though it has limited environments and agents now, the community is evolving.– Anurag PriyadarshiJun 2, 2016 at 10:22