How Learning Can Guide Evolution

By Geoffrey E. Hinton, Steven J. Nowlan. 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, methodology): “In our approach, we evaluate architectures without weight training, bypassing the costly inner loop, similar to the random trial approach in [44, 99] that evolved architectures to be more weight tolerant.”
Evolving Deep Neural Networks (2017, methodology): “Much of the work is based on utilizing the Baldwin effect, where learning only affects the selection [21].”
Born to Learn: the Inspiration, Progress, and Future of Evolved Plastic Artificial Neural Networks (2017, background): “The studies in Smith (1986); Hinton and Nowlan (1987); Boers et al. (1995); Mayley (1996); Bullinaria (2001) demonstrated the Baldwin effect with computational simulations.”
Imitative and Direct Learning as Interacting Factors in Life History Evolution (2017, unknown): “Belew [4] and Best [5] have already introduced imitation-based cultural factors into Hinton and Nowlanʼs [44] muchstudied model of learning guiding evolution, but their approaches are far removed from the neuralinspired learning relevant to the life history factors of interest here.”
Evolution of homing navigation in a real mobile robot (1996, result): “the selection procedure can evaluate a large set of values for each string (due to the oscillations of performance on the tness surface caused by synaptic change), rather than a single value [37] as in our case.”
Guiding or Hiding: Explorations into the Effects of Learning on the Rate of Evolution (1997, methodology): “It was rst tested empirically on drosophila by Waddington in the 1950's (Waddington , 1953, 1956) and was nally brought to the attention of the A-life/Adaptive Behaviour communities by Hinton and Nowlan (1987).”
Empirical Investigation of the Benefits of Partial Lamarckianism (1997, background): “Many researchers [1,4,16,20,27] have shown that GAs perform well for global searching because they are capable of quickly finding and exploiting promising regions of the search space, but they take a relatively long time to converge to a local optimum.”
The power of imperfect imitation (2013, background): “1 This model employs the basic structural form of Holland (1975), March (1991), and Bruderer and Singh (1996).”
Influence of Plasticity and Learning on Evolution under Directional Selection (2007, background): “An extreme case of the latter situation is the needle-in-a-haystack fitness landscape, assumed in simulation models that predicted an accelerating effect of both noise and learning on the rate of evolution (Hinton and Nowlan 1987; Fontanari and Meir 1990; French and Messinger 1994).”
Connectionist Learning Procedures (1989, background): “The evolutionary search for good architectures may actually be guided by learning [43].”
Adaptation of Robot Behaviour through Online Evolution and Neuromodulated Learning (2012, background): “There is much evidence that: (i) both processes are integral to the success of evolution in both biological and artificial systems [14, 19], and (ii) that learning can accelerate the evolution of good solutions [9], which is known as the Baldwin effect.”
Learning Evaluation Functions to Improve Optimization by Local Search (2000, methodology): “Some GA implementations do manage to take advantage of local search operators more sophisticated than bit- ips, using the trick of embedding a hillclimbing search into each objective function evaluation (Hinton & Nowlan, 1987).”
Adaptive versus non‐adaptive phenotypic plasticity and the potential for contemporary adaptation in new environments (2007, background): “Indeed, while a variety of models confirm that adaptive plasticity may facilitate adaptive evolution (e.g. Hinton & Nowlan 1987; Behera & Nanjundiah 2004; Ancel 1999), others have shown how plasticity slows the rate of evolution (e.g. Behera & Nanjundiah 1995; Ancel 2000, Huey, Hertz & Sinervo…”
Evolutionary ecology of learning: insights from fruit flies (2009, unknown): “This scenario, known as the Baldwin effect, has been demonstrated in several specific models (Hinton and Nowlan 1987; Fontanari and Meir 1990).”