An analysis of cooperative coevolutionary algorithms

By R. P. Wiegand, K. A. Jong. 2004

266 citations recorded by the source corpus.

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

Lenient Multi-Agent Deep Reinforcement Learning (2017, background): “It was designed to prevent relative overgeneralization [36], which occurs when agents gravitate towards a robust but suboptimal joint policy due to noise induced by the mutual influence of each agent’s exploration strategy on others’ learning updates.”
Evolutionary Dynamics of Multi-Agent Learning: A Survey (2015, background): “This problem is also known as relative overgeneralisation (Wiegand, 2003) or action shadowing (Fulda & Ventura, 2007).”
Accelerated Neural Evolution through Cooperatively Coevolved Synapses (2008, background): “…for using the cooperative coevolutionary approach is based on the intuition that many problems may be decomposable into weakly coupled low-dimensional subspaces that can be searched semi-independently by separate species (Wiegand et al., 2001; Jansen and Wiegand , 2003, 2004; Panait et al., 2006).”
An Overview of Cooperative and Competitive Multiagent Learning (2005, methodology): “The first one, Evolutionary Game Theory, EGT was successfully used to study the properties of cooperative coevolution [48, 195], to visualize basins of attraction to Nash equilibria for cooperative coevolution [121], and to study trajectories of concurrent Q-learning”
Biasing Coevolutionary Search for Optimal Multiagent Behaviors (2006, background): “The focus of this paper is on another prominent pathology that is called relative overgeneralization [24].”
Evolutionary computation: a unified approach (2018, background): “One promising approach currently being explored is the use of evolutionary game theory (Ficici and Pollack, 2000; Wiegand, 2004).”
Theoretical Advantages of Lenient Learners: An Evolutionary Game Theoretic Perspective (2008, background): “…coevolutionary algorithms Potter and Jong (1994); Potter (1997) can not only converge but also be attracted to joint suboptima in the search space (Wiegand, 2004), despite traditional evolutionary algorithms being guaranteed to converge to the global optimum when given enough time and sufficient…”
An Evolutionary Game Theoretic Analysis of Difference Evaluation Functions (2015, methodology): “EGT is well suited as a model of cooperative coevolutionary algorithms, and thus provides a theoretical framework in which we can analyze the effects of different fitness assignment operators [6, 9, 10, 11, 13, 14].”
Evolutionary computation and structural design: A survey of the state-of-the-art (2005, background): “Table 3: Coevolutionary architectures described in terms of attributes [188] No.”
A survey on optimization metaheuristics (2013, methodology): “Such a process can be static, in the sense that the division s for the separate componen ts are decided a priori and never altered, or dynamic, in the sense that populations of component s may be added or removed as the run progress es [279].”
Lenient Learning in Independent-Learner Stochastic Cooperative Games (2016, background): “Overcoming relative overgeneralization may require a more explorative approach than simple epsilon-greedy action selection: indeed as shown by Wiegand (2004), relative overgeneralization can cause players to not just hill-climb into local optima but get actively sucked into them despite action…”
Lenience towards Teammates Helps in Cooperative Multiagent Learning (2005, background): “Wiegand [16] argues that the sampling of collaborators introduces a certain bias on the search: wider peaks (see Figure 1) are more likely to have collaborators sampled from them, and this may bias the search toward suboptimal solutions; this is termed”
The road to everywhere : Evolution, complexity and progress in natural and artificial systems (2008, unknown): “[186]”