Learning to Adapt to Changing Environments in Evolving Neural Networks

By Stefano Nolfi, Domenico Parisi. 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

Born to Learn: the Inspiration, Progress, and Future of Evolved Plastic Artificial Neural Networks (2017, background): “This made sense as well as the intuitive concept that evolved plastic neural networks increase the adaptability to changing ER environments (Nolﬁ and Parisi, 1996).”
Learning and Evolution (1999, methodology): “If we allow our individuals to modify the genetically inherited standard weights but we randomize the teaching weights, in this case too learning will destroy whatever ability to explore is present at birth rather than increasing that ability (see Nolfi and Parisi, 1997).”
Evolving Mobile Robots in Simulated and Real Environments (1995, background): “Recently, Nolfi and Parisi [25, 26] showed how this approach can be generalized to a more complex task in which a Khepera with an arm and a gripper evolves an ability to recognize and pick up small round objects while avoiding walls.”
Evolving plastic neural networks with novelty search (2010, background): “For example, Nolfi and Parisi (1993, 1996) evolved self-teaching networks that trained a motor control network through backpropagation from the outputs of a teaching subnetwork.”
Neuroevolution: from architectures to learning (2008, background): “Nolfi and Parisi [56] evolved a neural controller for a mobile robot whose output layer included two ''teaching neurons'' that were used to modify the connection weights from the sensory neurons to the motor neurons with backpropagation learning during the robot's lifetime.”
Evolving adaptive neural networks with and without adaptive synapses (2003, background): “Nolfi and Parisi used backpropagation inside evolved networks to make predictions about future states (Nolfi et al. 1990, 1994) and also for a network to train itself ( Nolfi and Parisi 1993, 1995 ).”
Culling and Teaching in Neuro-Evolution (1997, background): “Nolfi and Parisi (1994, 1995) studied neuro-evolution in rapidly changing environments.”
The challenges of joint attention (2006, background): “However, such kinds of evolutionary approaches do not directly deal with the development of turn-taking behavior (but some researchers argue that these two forms of adaptive processes have complementary characteristics and can be effectively integrated ( Nolfi and Parisi, 1997; Floreano and Mondada”
Symbiotic Evolution of Neural Networks in Sequential Decision Tasks (1997, background): “Several researchers have investigated the use of local learning after each action in an evolutionary algorithm (Grefenstette 1991; Gruau and Whitley 1993; Littman 1995; Nol and Parisi 1995).”
Accelerated Neural Evolution through Cooperatively Coevolved Synapses (2008, methodology): “More recently, methods for evolving artificial neural networks or neuroevolution have shown promising results on continuous, partially observable tasks (Gomez, 2003; Nolfi and Parisi, 1995; Yamauchi and Beer, 1994).”
Fast learning for problem classes using knowledge based network initialization (2000, unknown): “Nolfi and Parisi (1997) and Sasaki and Tokoro (1997) have shown that in principle it is possible with such a method to generate individuals that can adapt faster to two different tasks.”
Evolutionary Computation for Reinforcement Learning (2012, background): “ents (McQuesten and Miikkulainen, 1997), trained them to predict state transitions (Nolfi et al, 1994), and trained them to teach themselves (Nolfi and Parisi, 1997).”
Structure learning in action (2010, background): “The applicability of the concept, however, stretches much farther and includes, for example, even ‘learning to learn’ on an evolutionary scale where our ontogenetic learning is merely an adaptation to a particular instantiation of a living environment for which we have been selected [101].”
Robust non-linear control through neuroevolution (2003, background): “Recently, significant progress has been made in solving continuous, non-Markov reinforcement learning problems using methods that evolve neural networks orneuroevolution (NE; Yamauchi and Beer 1994; Nolfi and Parisi 1995; Yao 1993; Moriarty 1997).”
Optimization for problem classes-neural networks that learn to learn (2000, background): “It has been shown experimentally that evolution of the initial weights in changing environments yields individuals with a higher ability to adapt to environmental changes (Nolfi, Miglino, and Parisi 1994; Nolfi and Parisi 1997; Sasaki and Tokoro 1997; Sasaki and Tokoro 1998).”