pbo
SpaceDrones
pbo | SpaceDrones | |
---|---|---|
1 | 1 | |
14 | 4 | |
- | - | |
3.0 | 7.8 | |
7 months ago | 8 months ago | |
Python | Python | |
MIT License | MIT License |
Stars - the number of stars that a project has on GitHub. Growth - month over month growth in stars.
Activity is a relative number indicating how actively a project is being developed. Recent commits have higher weight than older ones.
For example, an activity of 9.0 indicates that a project is amongst the top 10% of the most actively developed projects that we are tracking.
pbo
-
Why we use diagonal gaussian rather than multivariate guassian (with full covariance matrix)
I have implemented full covariance matrix output from neural networks in a drl-related project, in which I devised an optimization technique out of mixing a PG approach with CMA-ES concepts, and you can find a possible implementation of how to do so here: https://github.com/jviquerat/pbo In this specific context, full covariance matrix gave a very significant performance boost. Yet I don't want to draw premature conclusions on whether this will end up the same when plugged into PPO.
SpaceDrones
-
SpaceDrones: A simple learning environment for genetic optimization
Github: https://github.com/kaifishr/SpaceDrones
What are some alternatives?
vizier - Python-based research interface for blackbox and hyperparameter optimization, based on the internal Google Vizier Service.
RocketLander - A simple framework equipped with optimization algorithms, such as reinforcement learning, evolution strategies, genetic optimization, and simulated annealing, to enable an orbital rocket booster to land autonomously.
GeneticAlgorithmPython - Source code of PyGAD, a Python 3 library for building the genetic algorithm and training machine learning algorithms (Keras & PyTorch).
TorchGA - Train PyTorch Models using the Genetic Algorithm with PyGAD
evotorch - Advanced evolutionary computation library built directly on top of PyTorch, created at NNAISENSE.
zoofs - zoofs is a python library for performing feature selection using a variety of nature-inspired wrapper algorithms. The algorithms range from swarm-intelligence to physics-based to Evolutionary. It's easy to use , flexible and powerful tool to reduce your feature size.