Neuroevolution algorithms search optimal solutions by evolution rather than learning, which enables them to deal with large complex problems.
In this project, we are going to implement neuroevolution algorithms in a unified framework. Specifically, the algorithms we want to implement includes CNE, NEAT, HyperNEAT, etc.
Our goals include:
- Implementing neuroevolution algorithms for mlpack, include CNE, NEAT and HyperNEAT.
- Testing implemented algorithms by problems such as XOR problem, double pole balancing problem and so on.
- Writing corresponding document for implemented neuroevolution module.