Modified Selection and Search in Learning Automata Based Artificial Bee Colony in Noisy Environment

Rakshit, Pratyusha; Konar, Amit; Nagar, Atulya K.

doi:10.1109/cec.2019.8790092

Cited by 2 publications

(2 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The comparative framework is formed by considering a few widely popular noisy single objective optimization algorithms (NSOOAs), including learning automata induced noisy bee colony (LANBC) [21], memetic for uncertainties DE (MUDE) [22], subset based LA incorporated particle swarm optimization (LAPSO) [23], and noise tolerant genetic algorithm (NTGA) [24]. In the present work, the sampling strategy of extended LANBC [25] is replaced with the proposed TDQL induced adaptive sampling policy while keeping other strategies embedded in extended LANBC unchanged. The new contender algorithm thus developed is referred to as extended QLNBC henceforth and is regarded as new member of the comparative framework.…”

Section: B Comparative Frameworkmentioning

confidence: 99%

“…The performance of the proposed noisy single objective optimization algorithm, referred to as Qlearning induced noisy bee colony (QLNBC), is compared with four state-of-the-art techniques ‫. ]42ޤ12[‬ The proposed adaptive sampling strategy is also used to replace the sampling policy recommended in [25] and results in a new competitor in the comparative framework, called extended QLNBC. The comparative performance is analyzed with respect to function error value metric while optimizing noisy versions of 28 CEC'2013 benchmark functions [26].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Q-Learning Induced Artificial Bee Colony for Noisy Optimization

Rakshit

Konar

Nagar

2020

2020 IEEE Congress on Evolutionary Computation (CEC)

Self Cite

View full text Add to dashboard Cite

The paper proposes a novel approach to adaptive selection of sample size for a trial solution of an evolutionary algorithm when noise of unknown distribution contaminates the objective surface. The sample size of a solution here is adapted based on the noisy fitness profile in the local surrounding of the given solution. The fitness estimate and the fitness variance of a sub-population surrounding the given solution are jointly used to signify the degree of noise contamination in its local neighborhood (LN). The adaptation of sample size based on the characteristics of the fitness landscape in the LN of a solution is realized here with the temporal difference Q-learning (TDQL). The merit of the present work lies in utilizing the reward-penalty based reinforcement learning mechanism of TDQL for sample size adaptation. This sidesteps the prerequisite setting of any specific functional form of relationship between the sample size requirement of a solution and the noisy fitness profile in its LN. Experiments undertaken reveal that the proposed algorithms, realized with artificial bee colony, significantly outperform the existing counterparts and the state-of-the-art algorithms.

show abstract