Evolving large-scale neural networks for vision-based reinforcement learning
Explore this paper's citation graph
Summary
This paper scale-up their compressed network encoding where network weight matrices are represented indirectly as a set of Fourier-type coefficients, to tasks that require very-large networks due to the high-dimensionality of their input space.
- Type
- article
- Published
- 2013-07-06
- Cited by
- 172
- References
- 18
- OpenAlex
- https://openalex.org/W2038794597
- Semantic Scholar
- https://api.semanticscholar.org/CorpusID:3840949
Keywords
Reinforcement learning, Neuroevolution, Computer science, Artificial intelligence, Artificial neural network
References
- Designing Neural Networks Using Genetic Algorithms with Graph Generation System
- Intrinsically Motivated Evolutionary Search for Vision-Based Reinforcement Learning
- A Frequency-Domain Encoding for Neuroevolution
- Searching for Minimal Neural Networks in Fourier Space
- Generalized compressed network search
- Dynamic model of the octopus arm. I. Biomechanics of the octopus reaching movement.
- An adaptive 'broom balancer' with visual inputs
- Evolving neural networks in compressed weight space
- Machine Perception of Three-Dimensional Solids
- A novel generative encoding for exploiting neural network sensor and output geometry
- The 2009 Simulated Car Racing Championship
- Generating large-scale neural networks through discovering geometric regularities
- Accelerated Neural Evolution through Cooperatively Coevolved Synapses
- Discovering Neural Nets with Low Kolmogorov Complexity and High Generalization Capability
- Compressed Network Complexity Search
- , Ranit Aharonov , Yaakov Engel , Binyamin of the Octopus Reaching Movement Dynamic Model of the Octopus Arm
- I and J
Cited by
- Neuroevolution in Games: State of the Art and Open Challenges
- Do Artificial Reinforcement-Learning Agents Matter Morally?
- Language Understanding for Text-based Games using Deep Reinforcement Learning
- From Pixels to Torques: Policy Learning with Deep Dynamical Models
- Deep learning in neural networks: An overview
- Evolving deep unsupervised convolutional networks for vision-based reinforcement learning
- Multiagent cooperation and competition with deep reinforcement learning
- On Learning to Think: Algorithmic Information Theory for Novel Combinations of Reinforcement Learning Controllers and Recurrent Neural World Models
- MAIA: The role of innate behaviors when picking flowers in Minecraft with Q-learning
- Retaining Experience and Growing Solutions
- Exploratory Robotic Controllers : An Evolution and Information Theory Driven Approach. (Exploration Robotique Autonome hybridant : évolution et théorie de l'information)
- POET: An Evo-Devo Method to Optimize the Weights of Large Artificial Neural Networks
- Is depth information and optical flow helpful for visual control?
- Draft: Deep Learning in Neural Networks: An Overview
- Platform for Rapid Prototyping of AI Architectures
- Zero Shot Learning for Semantic Boundary Detection - How Far Can We Get?
- A Wavelet-based Encoding for Neuroevolution
- Learning a Driving Simulator
- Playing FPS Games with Deep Reinforcement Learning
- Extracting Cognition out of Images for the Purpose of Autonomous Driving
Related papers
- Incorporating Advice into Neuroevolution of Adaptive Agents
- Incorporating Advice into Evolution of Neural Networks
- A Neuroevolution Approach to General Atari Game Playing
- NEAT for large-scale reinforcement learning through evolutionary feature learning and policy gradient search
- Automated state feature learning for actor-critic reinforcement learning through NEAT
- Towards continual reinforcement learning through evolutionary meta-learning