Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Cool game. #2

Open
MariasStory opened this issue Sep 21, 2016 · 1 comment
Open

Cool game. #2

MariasStory opened this issue Sep 21, 2016 · 1 comment

Comments

@MariasStory
Copy link

Hi, cool game indeed. I am wondering:

  1. Can the trained network be saved and restored later?
  2. Can a deep network be used?
  3. Is it possible to put some additional obstacles on the road?
  4. Is it possible to decrease variation rate with longer path?
@TomaszRewak
Copy link
Owner

Hi, thanks.
It's still an early version, so I'm open to suggestions.
Right now I'm working on my engineering thesis, so I'm not planning on making any changes right now, but I will add these suggestions to my back list.
But in more details:

  1. Right now it's not possible. I actually don't think that it's so necessary, as the learning process is quite fast. But maybe in the future.
  2. You can define multiple hidden layers in the config. For example:
    { "inputs": 8, "hiddenLayers": [10,30,15] }
    It will define a network with 3 hidden layers with 10, 30 and 15 nodes respectively. You can also change the activation funstion for each node in "Activation function" section.
    You can teach them only by using genetic algorithm tho.
  3. It's not possible on the website, but it can be easly added by editing the source code. There are already redy components just wating to be used in scenarios like that.
  4. The map genarator is fixed within the source code. It can be altered there, but the website doesn't allow that (at this point).

I'm sorry for the long response time, but I was on a short vacations.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants