Filters








132,427 Hits in 6.1 sec

An autonomous explore/exploit strategy

Alex McMahon, Dan Scott, Will Browne
2005 Proceedings of the 2005 workshops on Genetic and evolutionary computation - GECCO '05  
Human beings manage this balance through logic and explorations based on feedback from the environment.  ...  The optimal balance between exploring and exploiting changes as the training progresses due to the increasing amount of learnt knowledge.  ...  are accurate and lead to a reward from every location, and so it stops HECS XCS HECS XCS exploring the environment and simply exploits its knowledge.  ... 
doi:10.1145/1102256.1102280 dblp:conf/gecco/McMahonSB05 fatcat:vuqkmebcmfgz3cqdhwq73rtkci

Functions for the analysis of exploration and exploitation

Daniel Ashlock, Nikola Krsic, Gary B. Fogel
2013 2013 IEEE Congress on Evolutionary Computation  
This paper suggests that representations themselves are just as important to consider when optimizing evolutionary algorithms for a balance of exploration and exploitation, and further that our system  ...  In this study we define a protocol to understand the exploratory and exploitative behaviors of evolutionary algorithms and evaluate two representations on four test functions.  ...  Previous Efforts Evaluating Exploration and Exploitation Significant effort in the literature has focused on the proper tradeoff of exploration and exploitation by evolutionary algorithms in optimization  ... 
doi:10.1109/cec.2013.6557807 dblp:conf/cec/AshlockKF13 fatcat:z3ruekkuvbdz3bpd6wqbkel2wa

Superior exploration-exploitation balance with quantum-inspired hadamard walks

Sisir Koppaka, Ashish Ranjan Hota
2010 Proceedings of the 12th annual conference comp on Genetic and evolutionary computation - GECCO '10  
The intuitive reasoning behind this approach, and the exploration-exploitation balance thus occurring is explained.  ...  From the results of the experiments carried out on the 0,1-knapsack problem, HQEA performs significantly better than a conventional genetic algorithm, CGA, and two quantum-inspired evolutionary algorithms  ...  By operating with a high n value on the worst solutions, exploration is achieved. By using a smaller n value on the good solutions, exploitation is achieved.  ... 
doi:10.1145/1830761.1830878 dblp:conf/gecco/KoppakaH10 fatcat:uzjpgdmc6rgohfzhlzmxklptze

Theoretical analysis of rank-based mutation - combining exploration and exploitation

Pietro S. Oliveto, Per Kristian Lehre, Frank Neumann
2009 2009 IEEE Congress on Evolutionary Computation  
Therefore it was proposed that the population be ranked according to fitness and the mutation rate of an individual should depend on its rank.  ...  To this end we theoretically explain the behaviour of rank-based mutation on various fitness landscapes proposed in the experimental work and present new significant classes of functions where the use  ...  Rowe for an interesting discussion on the use of rank-based variation operators.  ... 
doi:10.1109/cec.2009.4983114 dblp:conf/cec/OlivetoLN09 fatcat:meucrtm22faz3ggmoqfq7akzcq

Balancing the exploration and exploitation capabilities of the Differential Evolution Algorithm

M.G. Epitropakis, V.P. Plagianakos, M.N. Vrahatis
2008 2008 IEEE Congress on Evolutionary Computation (IEEE World Congress on Computational Intelligence)  
In this paper we propose a hybrid approach that combines Differential Evolution mutation operators in an attempt to balance their exploration and exploitation capabilities.  ...  The hybridization and composition of different Evolutionary Algorithms to improve the quality of the solutions and to accelerate execution is a common research practice.  ...  The above approach creates new classes of DE algorithms that depend on the selection of the explorative and exploitive mutation operators.  ... 
doi:10.1109/cec.2008.4631159 dblp:conf/cec/EpitropakisPV08 fatcat:pgfwkm37pfb5lhprpw3fm5n4ke

An exploration-exploitation compromise-based adaptive operator selection for local search

Nadarajen Veerapen, Jorge Maturana, Frédéric Saubion
2012 Proceedings of the fourteenth international conference on Genetic and evolutionary computation conference - GECCO '12  
In evolutionary algorithms, diversity is a key concept. We consider a related idea: the similarity between the candidate solution and the solutions in the search trajectory.  ...  The proposed methods are tested on the Quadratic Assignment Problem and Asymmetric Traveling Salesman Problem.Liens[1] http://okina.univ-angers.fr/publications?  ... 
doi:10.1145/2330163.2330340 dblp:conf/gecco/VeerapenMS12 fatcat:47h3g7meqfd7hbjgeoih4bg3ba

Static learning particle swarm optimization with enhanced exploration and exploitation using adaptive swarm size

Aditya Panda, Srijan Ghoshal, Amit Konar, Bonny Banerjee, Atulya K Nagar
2016 2016 IEEE Congress on Evolutionary Computation (CEC)  
The proposed Static Learning Particle Swarm Optimization with Enhanced Exploration and Exploitation using Adaptive Swarm Size (SLPSO) algorithm is tested on a set of shifted and rotated benchmark problems  ...  The first phase is assigned to focus solely on better exploration of the search space. The second phase focuses on better utilization of the explored information.  ...  Optimization through evolutionary algorithm solely depends on better exploration and exploitation of the search space.  ... 
doi:10.1109/cec.2016.7744016 dblp:conf/cec/PandaGKBN16 fatcat:k3m3aafz3vc7rlny7gwli7nkvu

Centric selection

David Simoncini, Sébastien Verel, Philippe Collard, Manuel Clergue
2009 Proceedings of the 11th Annual conference on Genetic and evolutionary computation - GECCO '09  
Performances on the quadratic assignment problem and NK-Landscapes put in evidence an optimal exploration / exploitation trade-off on both of the classes of problems.  ...  In this paper, we study the exploration / exploitation trade-off in cellular genetic algorithms.  ...  Introduction The exploration/exploitation trade-off is an important issue in evolutionary computation.  ... 
doi:10.1145/1569901.1570023 dblp:conf/gecco/SimonciniVCC09 fatcat:hdx3xsk2abeqlelad3w4crauu4

A unification of the prevalent views on exploitation, exploration, intensification and diversification

Francois Fagan, Jan H. Van Vuuren
2013 International Journal of Metaheuristics (IJMHeur)  
Six prevalent views on exploitation and exploration are identified in the literature, each expressing a different aspect of these notions.  ...  The consistency and meaningfulness of these views are substantiated by their deducibility from the proposed novel definitions of exploitation and exploration, based on the hypothetical construct of a probable  ...  Opinions expressed and conclusions arrived at, are those of the authors and are not necessarily to be attributed to the NRF.  ... 
doi:10.1504/ijmheur.2013.056407 fatcat:vjiilb22yvci3avo2xjjuy2iuq

A parameter control method of evolutionary algorithms using exploration and exploitation measures with a practical application for fitting Sovova's mass transfer model

Shih-Hsi Liu, Marjan Mernik, Dejan Hrnčič, Matej Črepinšek
2013 Applied Soft Computing  
Exploration and exploitation are omnipresent terms in evolutionary computation community that have been broadly utilized to explain how evolutionary algorithms perform search.  ...  However, only recently exploration and exploitation measures were presented in a quantitative way enabling to measure amounts of exploration and exploitation.  ...  With such, a number of exploration and exploitation measures can be defined based on the characteristics of exploration and exploitation trees.  ... 
doi:10.1016/j.asoc.2013.05.010 fatcat:x4323z4ctjgbrlcwnxdbuckby4

Balanced Explore-Exploit clustering based Distributed Evolutionary Algorithm for Multi-objective Optimisation

Mariem GZARA, Abdelbasset ESSABRI
2011 Studies in Informatics and Control  
In this paper, we present a new clustering-based parallel multi-objective evolutionary algorithm that balances between the two main concepts in metaheuristics, which are exploration and exploitation of  ...  The proposed algorithm is implemented and tested on several standard multi-objective test functions using a network of multiple computers.  ...  The algorithm executes iteratively and in turn an exploration step and exploitation step ( Figure 2 ).  ... 
doi:10.24846/v20i2y201102 fatcat:7bnfl5zz4nb3bmeb77x5xdsmi4

Selecting for Selection: Learning To Balance Adaptive and Diversifying Pressures in Evolutionary Search [article]

Kevin Frans, L.B. Soros, Olaf Witkowski
2021 arXiv   pre-print
Inspired by natural evolution, evolutionary search algorithms have proven remarkably capable due to their dual abilities to radiantly explore through diverse populations and to converge to adaptive pressures  ...  In deceptive or hard-to-search fitness landscapes, greedy selection often fails, thus it is critical that selection functions strike the correct balance between gradient-exploiting adaptation and exploratory  ...  Once trained, Sel4Sel discovers selective behaviors which dynamically balance between exploration and exploitation, enabling efficient adaptation on the tasks at hand.  ... 
arXiv:2106.09153v1 fatcat:eqi3axaawraazlzsj6y5ja3jz4

Multi-objective Binary Search Optimisation [chapter]

Evan J. Hughes
2003 Lecture Notes in Computer Science  
This paper introduces a new algorithm that aims to provide controllable exploration and exploitation of the decision space with a very limited number of function evaluations.  ...  The paper compares the performance of the algorithm to a typical evolutionary approach.  ...  As noted in most evolutionary algorithms, it is wise to always have a low level of exploration, even in the exploitation phase.  ... 
doi:10.1007/3-540-36970-8_8 fatcat:zjchwdaqivb4vka5bqnd62aaea

The evolution of knowledge creation online

Ruqin Ren
2015 Proceedings of the 11th International Symposium on Open Collaboration - OpenSym '15  
Based on the theory of self-organization system and evolution theory, the processes of variation and selection are explained in depth with examples observed on Wikipedia.  ...  The underlying theoretical assumption is that even though online communities feature very different characteristics than traditional organizations, the basic processes of trial-and-error learning in evolutionary  ...  March [10] studies the relation between exploration and exploitation in the context of organizations, and clarifies key concepts around this issue.  ... 
doi:10.1145/2788993.2791320 dblp:conf/wikis/Ren15 fatcat:hmvnx5x37bblpahov2fswshx74

Synergies between Evolutionary Algorithms and Reinforcement Learning

Madalina M. Drugan
2015 Proceedings of the Companion Publication of the 2015 on Genetic and Evolutionary Computation Conference - GECCO Companion '15  
environments that are modelled as Markov decision processes (MDP) by rewarding good actions and punishing bad actions • The exploration / exploitation dilemma • exploration -> try out actions of which  ...  metric •The Kullback-Leibler divergence regret metric • Exploitation/exploration trade-off: •Exploration: sample scalarization functions, and pull arms that might be unluckily identified as suboptimal  ...  heuristics should be independent of the type of heuristic it is applied on 63 Concluding remarks on reinforcement learning for evolutionary computation • Most EC algorithms use model free RL or multi-armed  ... 
doi:10.1145/2739482.2756582 dblp:conf/gecco/Drugan15 fatcat:5jedfs4jmfgclcpppzcjvz6yvu
« Previous Showing results 1 — 15 out of 132,427 results