Evolving Simple Organisms
nathanrooy.github.io
Evolving Simple Organisms
1–10 of 10 posts
Re: Evolving Simple Organisms
#2Correct me if I’m wrong, but doesn’t inheriting by averaging the parents’ matrices only work in single layer perceptrons?
Anyway, nice exercise! :)
Re: Evolving Simple Organisms
#3This is neat, if you want to extend it there's a version of genetic algorithm for neural networks that tries out different connection settings and hidden layer sizes -- it's called NEAT.
Re: Evolving Simple Organisms
#4Cool, I think it's Matt Stonie on the right in the competitive eating GIF.
Re: Evolving Simple Organisms
#5Reminds me of an old Javascript experiment called "Creatures Avoiding Planks" also based on neuroevolution:
Link: http://otoro.net/planks/ Past Discussion: https://news.ycombinator.com/item?id=10711951
Re: Evolving Simple Organisms
#6The article shows a useful neural network model for changing direction depending on an input. However, I'm not sure what this has to do with deep learning as there is only one hidden layer and even that one is only fully connected to output. Edit: also, why evolve weights when you can backpropagate?
Re: Evolving Simple Organisms
#7> ef evolve
Missing 'd'
Re: Evolving Simple Organisms
#8This tutorial is almost the same as this ~10 years old tutorial, only a little bit modernised: http://www.ai-junkie.com/ann/evolved/nnt1.html
Re: Evolving Simple Organisms
#9The article shows a useful neural network model for changing direction depending on an input. However, I'm not sure what this has to do with deep learning as there is only one hidden layer and even that one is only fully connected to output. Edit: also, why evolve weights when you can backpropagate?
Re: Evolving Simple Organisms
#10The article shows a useful neural network model for changing direction depending on an input. However, I'm not sure what this has to do with deep learning as there is only one hidden layer and even that one is only fully connected to output. Edit: also, why evolve weights when you can backpropagate?
Even random search (in the weight space) can outperform gradient-based deep reinforcement learning algorithms on some Atari games. Genetic algorithms are quite competitive. Paper: https://arxiv.org/abs/1712.06567v1