Cognitive systems based on adaptive algorithms

Holland, John H.; Reitman, Judith Spencer

doi:10.1145/1045343.1045373

Cited by 373 publications

(287 citation statements)

References 0 publications

Supporting

Mentioning

276

Contrasting

Unclassified

Order By: Relevance

“…It is derived from the classifier systems model proposed by [28]. Shown on Figure 1, this approach uses cooperation mechanisms within a single population.…”

Section: Selectionmentioning

confidence: 99%

Cooperative Coevolution for Agrifood Process Modeling

Barrière

Lutton

Wuillemin

et al. 2013

EVOLVE- A Bridge Between Probability, Set Oriented Numerics and Evolutionary Computation

View full text Add to dashboard Cite

On the contrary to classical schemes of evolutionary optimisations algorithms, single population Cooperative Co-evolution techniques (CCEAs, also called "Parisian" approaches) make it possible to represent the evolved solution as an aggregation of several individuals (or even as a whole population). In other words, each individual represents only a part of the solution. This scheme allows simulating the principles of Darwinian evolution in a more economic way, which results in gain in robustness and efficiency. The counterpart however is a more complex design phase. In this chapter, we detail the design of efficient CCEAs schemes on two applications related to the modeling of an industrial agri-food process. The experiments correspond to complex optimisations encountered in the modeling of a Camembert-cheese ripening process. Two problems are considered:• A deterministic modeling problem, phase prediction, for which a search for a closed form tree expression is performed using genetic programming (GP).• A Bayesian network structure estimation problem. The novelty of the proposed approach is based on the use of a two step process based on an intermediate representation called independence model. The search for an independence model is formulated as a complex optimisation problem, for which the CCEA scheme is particularly well suited. A Bayesian network is finally deduced using a deterministic algorithm, as a representative of the equivalence class figured by the independence model.

show abstract

“…It is derived from the classifier systems model proposed by [28]. Shown on Figure 1, this approach uses cooperation mechanisms within a single population.…”

Section: Selectionmentioning

confidence: 99%

Cooperative Coevolution for Agrifood Process Modeling

Barrière

Lutton

Wuillemin

et al. 2013

EVOLVE- A Bridge Between Probability, Set Oriented Numerics and Evolutionary Computation

View full text Add to dashboard Cite

show abstract

“…Methods explicitly addressing the division of labour include learning classifier systems (Holland and Reitman, 1978), co-operative coevolution (Potter, 1997), and evolutionary divide and conquer (Rosca, 1997), as well as techniques embedded in the modularity methods listed above. In all these methods, the same question re-occurs.…”

Section: Modularity and Credit Assignment In Artificial Evolutionmentioning

confidence: 99%

A computational model of symbiotic composition in evolutionary transitions

Watson

Pollack

2003

Biosystems

View full text Add to dashboard Cite

Several of the major transitions in evolutionary history, such as the symbiogenic origin of eukaryotes from prokaryotes, share the feature that existing entities became the components of composite entities at a higher-level of organization. This composition of pre-adapted extant entities into a new whole is a fundamentally different source of variation from the gradual accumulation of small random variations, and it has some interesting consequences for issues of evolvability. Intuitively, the pre-adaptation of sets of features in reproductively independent specialists suggests a form of 'divide and conquer' decomposition of the adaptive domain. Moreover, the compositions resulting from one level may become the components for compositions at the next level, thus scaling-up the variation mechanism. In this paper, we explore and develop these concepts using a simple abstract model of symbiotic composition to examine its impact on evolvability. To exemplify the adaptive capacity of the composition model, we employ a scale-invariant fitness landscape exhibiting significant ruggedness at all scales. Whilst innovation by mutation and by conventional evolutionary algorithms becomes increasingly more difficult as evolution continues in this landscape, innovation by composition is not impeded as it discovers and assembles component entities through successive hierarchical levels.

show abstract

“…Learning classifier systems (LCS) [15] are a machine learning system, situated in an environment that presents a problem that the system tries to learn and solve. The learning process takes place in discontinuous time intervals, where the system is presented a state and it gives an answer to it; depending on the answer there is a reward.…”

Section: Extended Classifier System (Xcs)mentioning

confidence: 99%

Co-evolutionary Agent Self-Organization for City Traffic Congestion Modeling

Hercog¹

2004

Genetic and Evolutionary Computation – GECCO 2004

View full text Add to dashboard Cite

Abstract. This paper introduces a model based on a multi-agent system that learns using the extended classifier system (MAXCS). The agents are informed in the news, after the weather forecast, the congestion levels of the main city avenues and they decide which avenue to use the following day. The results are encouraging, as the agents, both homogeneous and heterogeneous adapt to the congestion thresholds set by the authorities. The main factors for this adaptation are the reward received by the agents and their perception.

show abstract

Cognitive systems based on adaptive algorithms

Cited by 373 publications

References 0 publications

Cooperative Coevolution for Agrifood Process Modeling

Cooperative Coevolution for Agrifood Process Modeling

A computational model of symbiotic composition in evolutionary transitions

Co-evolutionary Agent Self-Organization for City Traffic Congestion Modeling

Contact Info

Product

Resources

About