Sequential Behavior and Learning in Evolved Dynamical Neural Networks

By Brian Yamauchi, Randall D. Beer. 1994

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

Deep learning in neural networks: An overview (2014, background): “Many RNN evolvers have been proposed (e.g., Miller et al., 1989; Wieland, 1991; Cliff et al., 1993; Yao, 1993; Nolfi et al., 1994a; Sims, 1994; Yamauchi and Beer, 1994; Miglino et al., 1995; Moriarty, 1997; Pasemann et al., 1999; Juang, 2004).”
On Learning to Think: Algorithmic Information Theory for Novel Combinations of Reinforcement Learning Controllers and Recurrent Neural World Models (2015, methodology): “1 RL through Direct and Indirect Search in RNN Program Space It is possible to train small RNNs with a few 100 or 1000 weights using evolutionary algorithms [200, 255, 105, 56, 68] to search the space of NN weights [165, 307, 44, 321, 180, 259, 320, 164, 173, 69, 71, 187, 121, 313, 66, 270, 269, 305”
Exploring the T-Maze: Evolving Learning-Like Robot Behaviors Using CTRNNs (2003, methodology): “As in [12] the Forward Euler numerical integration method with step size ∆t = 1 is applied to equation (1).”
Evolution of an artificial neural network based autonomous land vehicle controller (1996, methodology): “Previous studies which have used evolution as the principal learning paradigm for training artificial neural networks have often modeled evolution through genetic algorithms [5], [61, [9], [201, [221, [29], [ 31 ].”
Learning to perceive the world as articulated: an approach for hierarchical learning in sensory-motor systems (1998, methodology): “The current study was inspired by other adaptive agent research based on the dynamical systems approach by, e.g. Beer (1995), Yamauchi and Beer (1994) and Smithers (1996).”
Genetic Representation and Evolvability of Modular Neural Controllers (2010, background): “, [54], [55]), it is generally believed that synaptic plasticity is one of the basic principles that allows for learning and memory in biological nervous systems [56].”
Neuroevolution: from architectures to learning (2008, background): “For example, Yamauchi and Beer [90] evolved CTRNNs capable of generating and learning short sequences without synaptic plasticity.”
Evolution of Plastic Control Networks (2001, methodology): “…to other methodologies that exploit evolution for generating dynamical controllers that display adaptive behavior without using synaptic plasticity (Yamauchi and Beer, 1995) and for discovering the rules by which the logic functions of FPGA controllers change according to environmental situations…”
Evolutionary and Computational Advantages of Neuromodulated Plasticity (2008, background): “, 2007), there are examples where adaptive behaviour is achieved with fixed weight networks (Yamauchi and Beer, 1994; Stanley and Miikkulainen, 2003a).”
Evolutionary Robotics: A New Scientific Tool for Studying Cognition (2005, background): “Apart from this, the experimental regime follows very closely that of Yamauchi and Beer, except that the agent is started towards one side of the arena (randomly chosen), and it is only once it has reached the central portion that the target is placed randomly towards one end or the other of the are”
Levels of dynamics and adaptive behavior in evolutionary neural controllers (2002, methodology): “Two classes of dynamical recurrent neural networks, Continuous Time Recurrent Neural Networks (CTRNNs) (Yamauchi and Beer, 1994) and Plastic Neural Networks (PNNs) (Floreano and Urzelai, 2000) are compared on two behavioral tasks aimed at exploring their capabilities to display reinforcement-learnin”
Training Recurrent Networks by Evolino (2007, methodology): “more efficient and principled way is to search the space of RNN weight matrices [20,21,26,45,53,54] using evolutionary algorithms [1 3,31,41].”
Evolving spike-timing-dependent plasticity for single-trial learning in robots (2003, background): “Work on evolutionary robotics has begun to challenge this stationary perspective on learning by providing examples of evolved learning behaviour in robots using neural networks that lack synaptic plasticity, and yet retain enough dynamic plasticity to respond with di® erent patterns of activity in a”