Masaryk University, Brno, Czech Republic , Wed 08 Apr 2009

Slide1 l.jpg
1 / 74
0
0
750 days ago, 212 views
PowerPoint PPT Presentation

Presentation Transcript

Slide 1

The Bees Algorithm, and Its Applications Dr. Ziad Salem, HDD, PhD, BsC. Spezs1@hotmail.com Computer Engineering Department Electrical and Electronic Engineering Faculty Aleppo University, Aleppo, Syrian Arab Republic Masaryk University, Brno, Czech Republic , Wed 08 Apr 2009

Slide 2

Outlines Introduction Intelligent Swarm –based streamlining The Bees Algorithm Bees in Nature Proposed Bees Algorithm Simple Example BA Applications to Data mining Conclusion Masaryk University, Brno, Czech Republic , Wed 08 Apr 2009

Slide 3

Introduction There was an awesome enthusiasm between scientists to create look calculations that find close ideal arrangements in sensible running time The Swarm-based Algorithm is a hunt calculation fit for finding great arrangements effectively The calculation could be considered as having a place with the class of "Savvy Optimization Tools" Masaryk University, Brno, Czech Republic , Wed 08 Apr 2009

Slide 4

Swarm-based Optimization Algorithm SOAs mince nature's strategies to infer a pursuit towards the ideal arrangement The key distinction amongst SOAs and direct inquiry calculations, for example, Hill Climbing is that SOAs utilize a populace of answers for each emphasis rather than a solitary arrangement As a populace of arrangements is prepared in a cycle, the result of every cycle is additionally a populace of arrangements Masaryk University, Brno, Czech Republic , Wed 08 Apr 2009

Slide 5

H sick Climbing Is a numerical advancement procedure which has a place with the group of neighborhood inquiry It can be utilized to take care of issues that have numerous arrangements, some of which are superior to anything others It begins with an arbitrary (possibly poor) arrangement, and iteratively rolls out little improvements to the arrangement, every time enhancing it a bit. At the point when the calculation can't see any change any longer, it ends. In a perfect world, by then the present arrangement is near ideal, however it is not ensured that slope climbing will ever approach the ideal arrangement Masaryk University, Brno, Czech Republic , Wed 08 Apr 2009

Slide 6

H sick Climbing It can be connected to the voyaging Salesman Problem . It is anything but difficult to discover an answer that visits every one of the urban areas yet will be extremely poor contrasted with the ideal arrangement Hill climbing is utilized generally as a part of AI, for achieving an objective state from a beginning hub. Masaryk University, Brno, Czech Republic , Wed 08 Apr 2009

Slide 7

Swarm-based Optimization Algorithm If a streamlining issue has a solitary ideal, SOA populace individuals can be relied upon to join to that ideal arrangement If an advancement issue has numerous ideal arrangements, SOA can be utilized to catch them in its last populaces Masaryk University, Brno, Czech Republic , Wed 08 Apr 2009

Slide 8

Swarm-based Optimization Algorithm SOAs include: The Ant Colony Optimization ( ACO ) calculation The Genetic Algorithm ( GA ) The Particle Swarm Optimization ( PSO ) calculation Others… … ( Bees Algorithm ( BA )) Masaryk University, Brno, Czech Republic , Wed 08 Apr 2009

Slide 9

Ant Colony Optimization (ACO) ACO is an exceptionally fruitful calculation which imitates the conduct of genuine Ants are equipped for finding the most brief way from the nourishment source to their home utilizing a compound substance called pheromone to control their inquiry A passing lost subterranean insect will take after this trail relies on upon the nature of the pheromone laid on the ground as the ants move Masaryk University, Brno, Czech Republic , Wed 08 Apr 2009

Slide 10

Particle Swarm Optimization (PSO) PSO is an enhancement methodology in light of the social conduct of gatherings of associations (for instance the rushing of winged creatures and the tutoring of fish) Individual arrangements in a populace are seen as "particles" that develop or change their positions with time Each molecule adjusts its position in hunt space as indicated by its own particular experience furthermore that of a neighboring molecule by recalling that best position went by without anyone else's input and its neighbors (consolidating nearby and worldwide pursuit strategies) Masaryk University, Brno, Czech Republic , Wed 08 Apr 2009

Slide 11

Genetic Algorithm (GA) GA depends on common determination and hereditary recombination GA works by picking arrangements from the flow populace and after that apply hereditary administrators, (for example, transformation and hybrid ) to make another populace GA abuses authentic data to guess on new pursuit regions with enhanced execution GA advantage: It performs worldwide hunt Masaryk University, Brno, Czech Republic , Wed 08 Apr 2009

Slide 12

SOAs Applications SOA procedures can be utilized as a part of various applications The U.S. military is exploring swarm systems for controlling vehicles The European Space Agency is contemplating an orbital swarm for self get together NASA is examining the utilization of swarm innovation for planetary mapping Masaryk University, Brno, Czech Republic , Wed 08 Apr 2009

Slide 13

Application to SOAs in Data Mining Some analysts proposed a Particle Swarm Optimizer as a device for Data Mining They found that Particle Swarm Optimizers turned out to be an appropriate contender for arrangement undertakings Reference Tiago Sousa, Ana Neves, Arlindo Silva , Swarm Optimization as a New Tool for Data Mining . Procedures of the seventeenth International Symposium on Parallel and Distributed Processing , Page: 144.2   , 2003, ISBN:0-7695-1926-1 Masaryk University, Brno, Czech Republic , Wed 08 Apr 2009

Slide 14

The Bees Algorithm (BA) Bees in nature Proposed Bees Algorithm Masaryk University, Brno, Czech Republic , Wed 08 Apr 2009

Slide 15

Bees in Nature 1-A province of bumble bees can broaden itself over long separations in various bearings (more than 10 km) Masaryk University, Brno, Czech Republic , Wed 08 Apr 2009

Slide 16

Bees in Nature Flower patches with ample measures of nectar or dust that can be gathered with less exertion ought to be gone by more honey bees, while patches with less nectar or dust ought to get less honey bees Masaryk University, Brno, Czech Republic , Wed 08 Apr 2009

Slide 17

Bees in Nature 2-Scout honey bees seek arbitrarily starting with one fix then onto the next Masaryk University, Brno, Czech Republic , Wed 08 Apr 2009

Slide 18

Bees in Nature 3-The honey bees who come back to the hive, assess the diverse patches relying upon certain quality edge (measured as a blend of a few components, for example, sugar content) Masaryk University, Brno, Czech Republic , Wed 08 Apr 2009

Slide 19

Bees in Nature 4-They store their nectar or dust go to the " move floor " to play out a " waggle move " Masaryk University, Brno, Czech Republic , Wed 08 Apr 2009

Slide 20

Bees in Nature 5-Bees convey through this waggle move which contains the accompanying data: The course of blossom patches (point between the sun and the fix) The separation from the hive (length of the move) The quality rating (wellness) (recurrence of the move) Masaryk University, Brno, Czech Republic , Wed 08 Apr 2009

Slide 21

Bees in Nature These data helps the settlement to send its honey bees decisively 6-Follower honey bees pursue the artist honey bee to the fix to assemble sustenance productively and rapidly Masaryk University, Brno, Czech Republic , Wed 08 Apr 2009

Slide 22

Bees in Nature 7 a similar fix will be promoted in the waggle move again when coming back to the hive is it still adequate as a nourishment source ( relying upon the sustenance level ) and more honey bees will be selected to that source 8-More honey bees visit bloom patches with copious measures of nectar or dust Masaryk University, Brno, Czech Republic , Wed 08 Apr 2009

Slide 23

Bees in Nature Thus, as indicated by the wellness, patches can be gone by more honey bees or might be surrendered Masaryk University, Brno, Czech Republic , Wed 08 Apr 2009

Slide 24

Proposed Bees Algorithm (BA) The Bees Algorithm is an improvement calculation enlivened by the common searching conduct of bumble bees to locate the ideal arrangement The accompanying figure demonstrates the pseudo code of the calculation in its most straightforward frame Masaryk University, Brno, Czech Republic , Wed 08 Apr 2009

Slide 25

Proposed Bees Algorithm (BA) 1. Initialise populace with irregular arrangements. 2. Assess wellness of the populace. 3. While (ceasing rule not met)/Forming new populace. 4. Select locales for neighborhood look. 5. Enroll honey bees for chose destinations (more honey bees for best e locales) and assess fitnesses. 6. Select the fittest honey bee from every fix. 7. Dole out outstanding honey bees to seek arbitrarily and assess their fitnesses. 8. End While. Pseudo code of the fundamental honey bees calculation Masaryk University, Brno, Czech Republic , Wed 08 Apr 2009

Slide 26

Proposed Bees Algorithm (BA) 100 The calculation requires various parameters to be set: Number of scout honey bees n Number of destinations chose m out of n went by locales Number of best locales e out of m chose locales Number of honey bees enrolled for best e locales nep or ( n2 ) Number of honey bees enlisted for the other ( m-e ) selected locales which is nsp or ( n1 ) Initial size of patches ngh which incorporates site and its neighborhood and halting rule Number of calculation steps redundancies imax 10 3 40 in neighborhood region Rich Poor 20 0-1 (0.2) 10,300,1000 Masaryk University, Brno, Czech Republic , Wed 08 Apr 2009

Slide 27

Proposed Bees Algorithm (BA) The accompanying figure demonstrates the flowchart of the Basic Bees Algorithm Masaryk University, Brno, Czech Republic , Wed 08 Apr 2009

Slide 28

Initialise a Population of n Scout Bees Evaluate the Fitness of the Population Select m Sites for Neighborhood Search Determine the Size of Neighborhood (Patch Size ngh ) Neighborhood Search Recruit Bees for Selected Sites (more Bees for the Best e Sites) Select the Fittest Bee from Each Site Assign the ( n–m ) Remain

SPONSORS