A comparison between cellular encoding and direct encoding for genetic neural networks

By Frédéric Gruau, L. D. Whitley, Larry D. Pyeatt. 1996

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Weight Agnostic Neural Networks (2019, background): “Our work has connections to existing work not only in deep learning, but also to various other ﬁelds: Architecture Search Search algorithms for neural network topologies originated from the ﬁeld of evolutionary computing in the 1990s [1, 8, 17, 25, 33, 40, 53, 59, 70, 71, 81, 86, 116, 117].”
Deep learning in neural networks: An overview (2014, background): “Instead of evolving large NNs directly (Section 6.6), one can sometimes greatly reduce the search space by evolving compact encodings of NNs, e.g., through Lindenmeyer Systems (Jacob, Lindenmayer, & Rozenberg, 1994; Lindenmayer, 1968), graph rewriting (Kitano, 1990), Cellular Encoding (Gruau, Whitle”
On Learning to Think: Algorithmic Information Theory for Novel Combinations of Reinforcement Learning Controllers and Recurrent Neural World Models (2015, methodology): “However, the search space can often be reduced dramatically by evolving compact encodings of neural networks (NNs), e.g., through Lindenmeyer Systems [115], graph rewriting [127], Cellular Encoding [83], HyperNEAT [268], and other techniques [245, Sec.”
Picbreeder: A Case Study in Collaborative Evolutionary Exploration of Design Space (2011, background): “It is now widely recognized that a good encoding allows information in the genotype to be reused in producing the phenotype (Gruau et al., 1996; Hart et al., 1994).”
Compositional Pattern Producing Networks : A Novel Abstraction of Development (2007, background): “neuroevolutionmethods, evolved either fixed-topology networks [30, 66], or arbitrary random-topology networks [5, 32, 82], NEAT is the first to begin evolution with a population of small, simple networks and complexifythe network topology over generations, leading to increasingly sophisticated behav”
Cartesian genetic programming encoded artificial neural networks: a comparison using three benchmarks (2013, background): “A restriction placed on the controller, also used by [6, 8, 45], is to limit the magnitude of the applied force to always be greater than 1 256 × 10N ; this ensures the controller cannot position the poles to be almost at the point of balance and then do nothing until they fall.”
Efficient Reinforcement Learning Through Evolving Neural Network Topologies (2002, methodology): “For example, on the most difficult versions of the pole balancing problem, which is the standard benchmark for reinforcement learning systems, NE methods have recently outperformed other reinforcement learning techniques (Gruau et al. 1996; Moriarty and Miikkulainen 1996).”
Evolving plastic neural networks with novelty search (2010, background): “Third, many systems that evolve network topologies and weights begin evolution with a population of random topologies (Gruau et al., 1996; Yao, 1999).”
Gaussian Processes for Regression and Optimisation (2007, background): “To prevent this, Gruau [22] defined a fitness function that penalises such solutions.”
Evolving a Roving Eye for Go (2004, background): “Third, unlike other systems that evolve network topologies and weights [17, 18] NEAT begins with a uniform population of simple networks with no hidden nodes.”
Compositional pattern producing networks: A novel abstraction of development (2007, background): “Both approaches have produced significant results, including simulated threedimensional body morphologies and locomotion [7, 15], neural networks with repeating structures [21, 27], and physical structures with self-similarity [7, 17].”
Single-unit pattern generators for quadruped locomotion (2013, background): “The simultaneous discovery of regularities is one reason indirect encodings [2, 5, 13, 16, 22, 32] have become an important area of research.”
Evolving Soccer Keepaway Players Through Task Decomposition (2005, methodology): “Neuroevolution has repeatedly been shown to be an effective reinforcement learning method for non-linear control tasks (Gruau et al., 1996; Moriarty and Miikkulainen, 1996; Stanley and Miikkulainen, 2004; Gomez, 2003).”
Evolutionary Function Approximation for Reinforcement Learning (2006, methodology): “In this thesis, I focus on using evolutionary methods [9] because of their demonstrated ability to discover effective representations [11, 21].”
The road to everywhere : Evolution, complexity and progress in natural and artificial systems (2008, unknown): “as double pole balancing ([76]).”
A Taxonomy for Artificial Embryogeny (2003, methodology): “[35] used grammar trees to encode steps in the development of a neural network starting from a single ancestor cell.”
A Hypercube-Based Encoding for Evolving Large-Scale Neural Networks (2009, background): “, neuroevolution methods), which traditionally evolve either fixed-topology networks [18, 45] or arbitrary random-topology networks [3, 22, 60].”