Interactions between Learning and Evolution : The Outstanding Strategy Generated by the Baldwin E ect

By Takaya Arita, Reiji Suzuki. 2000

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): “e, ﬁne-tune, and transfer skills throughout their lifespan, as in animals [125]. Inspired by the Baldwin effect [3], weight tolerant networks have long linked theories of evolution and learning in AI [1,48,106]. To develop a single WANN capable of encoding many different useful tasks in its environm”
Human-level performance in 3D multiplayer games with population-based reinforcement learning (2018, methodology): “IEEE Transactions on Autonomous Mental Development, 2(2):70–82, 2010.”
Culling and Teaching in Neuro-Evolution (1997, unknown): “For example, (Braun and Zagorski 1994; Sebag and Schoenauer 1994; Ackley and Littman 1991).”
Behavioural plasticity: an interaction between evolution and experience (2010, methodology): “A similar prediction has been obtained in artificial intelligence models (Belew 1989; Ackley and Littman 1990; Mayley 1997).”
Intrinsically Motivated Reinforcement Learning: An Evolutionary Perspective (2010, background): “Other researchers have reported interesting results of computational experiments involving evolutionary search for RL reward functions [1], [8], [19], [31], [43], but they did not directly address the motivational issues on which we focus.”
Evolutionary robotics: coping with environmental change (2000, methodology): “This is typically achieved by evolving neural controllers that learn with an oothe-shelf algorithm, such as reinforcement learning or back-propagation, starting from synaptic weights speciied on the genetic string of the individual (Ackley & Littman, 1992; Noll, Miglino, & Parisi, 1994).”
Evolutionary Computation for Reinforcement Learning (2012, methodology): “However, perhaps the most natural hybrids for the reinforcement learning setting are combination of evolution with temporal-difference methods (Ackley and Littman, 1991; Wilson, 1995; Downing, 2001; Whiteson and Stone, 2006a).”
Evolution and Learning in an Intrinsically Motivated Reinforcement Learning Robot (2007, background): “Because of these time-scale differences, the models proposed so far, which typically used neural networks as agents’ control systems, assigned to evolution the role of developing the ‘general aspects’ of learning systems, for example their overall architecture [6], the learning rules [7][8], the par”
Evolution-Guided Policy Gradient in Reinforcement Learning (2018, background): “From an evolutionary perspective, combining RL with EA is closely related to the idea of incorporating learning with evolution [1, 12, 57].”
Neuroevolution (2010, background): “networks that learn well will be selected for reproduction even if the weight changes themselves are not inherited [1, 3, 8].”
Phenotypic plasticity and evolution by genetic assimilation (2006, methodology): “Other authors have taken a network modeling approach (Masel, 2004), or used the prisoner’s dilemma framework (Suzuki and Arita, 2004).”
Learning and Evolution (1999, background): “38”