we wanna enrich our silly dog to behave like this:
- if you pet him, he sits down
- if you whistle, he barks
- if you whistle and pet him, he rolls
two inputs: PET, WHISTLE
three outputs: SIT, BARK, ROLL
You can adjust inputs and see the given percentage to what the output should look like, but also change the training iteration to see how it increases/decreseases in accuracy.
made with a little bit of help from Synaptic Library and this article