‫‪ K‬ﻤﺠﻠﺔ ﺍﻟﺘﺭﺒﻴﺔ ﻭﺍﻟﻌﻠﻡ ‪ -‬ﺍﻟﻤﺠﻠﺩ )‪ (19‬ﺍﻟﻌﺩﺩ )‪ (4‬ﻟﺴﻨﺔ ‪J 2007‬‬

‫ﺍﺴﺘﺨﺩﺍﻡ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﺍﻟﺠﻴﻨﻴﺔ ﻓﻲ ﺘﻁﺎﺒﻕ ﺃﻨﻤﺎﻁ ﺍﻟﺤﺭﻑ ﺍﻹﻨﻜﻠﻴﺯﻱ‬

‫ﻤﻬﺎ ﻋﺒﺩ ﺍﻹﻟﻪ ﻤﺤﻤﺩ ﺍﻟﺒﺩﺭﺍﻨﻲ‬

‫ﻗﺴﻡ ﺍﻟﺒﺭﻤﺠﻴﺎﺕ‪/‬ﻜﻠﻴﺔ ﻋﻠﻭﻡ ﺍﻟﺤﺎﺴﺒﺎﺕ ﻭﺍﻟﺭﻴﺎﻀﻴﺎﺕ‬
‫ﺠﺎﻤﻌﺔ ﺍﻟﻤﻭﺼل‬
‫ﺍﻟﻘﺒﻭل‬ ‫ﺍﻻﺴﺘﻼﻡ‬
‫‪09/01/2007‬‬ ‫‪02/08/2006‬‬
‫‪ABSTRACT‬‬
‫‪The focus of the present research is on the issue of compatibility of‬‬
‫‪the patterns of the English letter through the use of a probable research‬‬
‫‪Algorithm called the Genetic Algorithm.‬‬
‫‪The research clarifies the ability of the algorithm in the compatibility‬‬
‫‪of patterns between the original(ideal) pattern of the letter and the‬‬
‫‪deformed patterns since the genetic algorithm is good for some tasks that‬‬
‫‪require examples. It applies to the issues that have large (wide)areas and‬‬
‫‪large variables. It can also be easily in addition quickly solved, in‬‬
‫‪addition to that ,it provides a solution that is quite near to the ideal‬‬
‫‪solution of the patterns used. The results have proved the ability of the‬‬
‫‪algorithm to carry out the process of the compatibility of the patterns of‬‬
‫‪English Letters compared with other algorithms.‬‬

‫ﺍﻟﻤﻠﺨﺹ‬
‫ﺘﺭﻜﺯ ﺍﻻﻫﺘﻤﺎﻡ ﻓﻲ ﻫﺫﺍ ﺍﻟﺒﺤﺙ ﻋﻠﻰ ﻤﺴﺄﻟﺔ ﺘﻁﺎﺒﻕ ﺍﻷﻨﻤﺎﻁ ﻟﻠﺤﺭﻑ ﺍﻹﻨﻜﻠﻴـﺯﻱ ﺒﺎﺴـﺘﺨﺩﺍﻡ‬
‫ﺨﻭﺍﺭﺯﻤﻴﺔ ﺒﺤﺙ ﺍﺤﺘﻤﺎﻟﻴﺔ ﺘﺩﻋﻰ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﺍﻟﺠﻴﻨﻴﺔ‪.‬‬
‫ﻴﻭﻀﺢ ﺍﻟﺒﺤﺙ ﺇﻤﻜﺎﻨﻴﺔ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﻓﻲ ﺘﻁﺎﺒﻕ ﺍﻷﻨﻤﺎﻁ ﻤﺎ ﺒﻴﻥ ﺍﻟﻨﻤﻁ ﺍﻷﺼﻴل )ﺍﻟﻤﺜـﺎﻟﻲ(‬
‫ﻟﻠﺤﺭﻑ ﻭﺍﻷﻨﻤﺎﻁ ﺍﻟﻤﺸﻭﻫﺔ‪ ،‬ﻟﻜﻭﻥ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﺍﻟﺠﻴﻨﻴﺔ ﺠﻴﺩﺓ ﻟﻌﺩﺩ ﻤﻥ ﺍﻷﻋﻤـﺎل ﺍﻟﺘـﻲ ﺘﺘﻁﻠـﺏ‬
‫ﺍﻷﻤﺜﻠﺔ‪ ،‬ﻓﻬﻲ ﺘﻁﺒﻕ ﻋﻠﻰ ﺍﻟﻤﺴﺎﺌل ﺍﻟﺘﻲ ﺘﻤﺘﻠﻙ ﻤﺴﺎﺤﺔ ﻭﺍﺴﻌﺔ ﻭﻤﺘﻐﻴﺭﺍﺕ ﻜﺒﻴﺭﺓ ﻭﻓﻲ ﺍﻻﻤﻜﺎﻥ ﺤﻠﻬﺎ‬
‫ﺒﺴﻬﻭﻟﺔ ﻭﺴﺭﻋﺔ‪ ،‬ﻜﻤﺎ ﺇﻨﻬﺎ ﺘﻌﻁﻲ ﺤﻼ ﻗﺭﻴﺒﹰﺎ ﺠﺩﹰﺍ ﻤﻥ ﺍﻟﺤل ﺍﻟﻤﺜﺎﻟﻲ ﻟﻸﻨﻤﺎﻁ ﺍﻟﻤﺴﺘﺨﺩﻤﺔ‪ .‬ﺒﺭﻫﻨـﺕ‬
‫ﺍﻟﻨﺘﺎﺌﺞ ﻋﻠﻰ ﻗﺩﺭﺓ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﻋﻠﻰ ﺇﺠﺭﺍﺀ ﻋﻤﻠﻴﺔ ﺘﻁﺎﺒﻕ ﺍﻷﻨﻤﺎﻁ ﻟﻠﺤﺭﻭﻑ ﺍﻹﻨﻜﻠﻴﺯﻴـﺔ ﻤﻘﺎﺭﻨـﺔ‬
‫ﺒﺨﻭﺍﺭﺯﻤﻴﺎﺕ ﺃﺨﺭﻯ ‪.‬‬

‫‪ .1‬ﺍﻟﻤﻘﺩﻤﺔ‬
‫ﺇﻥ ﺍﺯﺩﻴﺎﺩ ﺍﻟﺼﻌﻭﺒﺎﺕ ﺍﻟﺘﻘﺎﻨﻴﺔ ﻭﺘﻌﻘﻴﺩﻫﺎ ﻭﻋﺩﻡ ﻗﺩﺭﺓ ﺍﻟﺤﻠﻭل ﺍﻟﺒﺭﻤﺠﻴﺔ ﺍﻟﺘﻘﻠﻴﺩﻴـﺔ ﻋﻠـﻰ‬
‫ﺍﺴﺘﻴﻌﺎﺒﻬﺎ ﻗﺎﺩ ﺍﻟﺒﺎﺤﺜﻴﻥ ﺇﻟﻰ ﺍﻟﺘﻌﻤﻕ ﻓﻲ ﺇﻴﺠﺎﺩ ﺨﻭﺍﺭﺯﻤﻴﺎﺕ ﻜﻔﻭﺀﺓ ﺘﺴﺎﻋﺩ ﻓـﻲ ﺇﻴﺠـﺎﺩ ﺍﻟﺤﻠـﻭل‬

‫‪84‬‬
 ‫ﺍﺴﺘﺨﺩﺍﻡ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﺍﻟﺠﻴﻨﻴﺔ ﻓﻲ ﺘﻁﺎﺒﻕ ‪....‬‬

‫ﺍﻟﻤﻨﺎﺴﺒﺔ ﻭﺍﻟﻤﺜﺎﻟﻴﺔ ﻟﻠﻤﺴﺎﺌل ﺍﻟﻤﻌﻘﺩﺓ ﻭﻓﻲ ﺴﺭﻋﺔ ﺍﻟﻭﺼﻭل ﺇﻟﻰ ﺍﻟﺤﻠﻭل ﻭﺨﺯﻨﻬﺎ ﻭﺍﺴﺘﺭﺠﺎﻋﻬﺎ ‪ ،‬ﺜﻡ‬
‫ﺍﻟﺘﻭﺼل ﺇﻟﻰ ﻫﻴﻜﻠﻴﺔ ﻤﺘﺭﺍﺒﻁﺔ ﻟﺒﻨﻴﺔ ﺒﺭﻤﺠﻴﺔ ﺫﻜﻴﺔ ﻭﻤﻥ ﻫﺫﻩ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺎﺕ ﺍﻟﻜﻔﻭﺀﺓ ﺍﻟﺨﻭﺍﺭﺯﻤﻴـﺔ‬
‫ﺍﻟﺠﻴﻨﻴﺔ )‪ ، (Genetic Algorithms‬ﻭﺘﻌﺩ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﺍﻟﺠﻴﻨﻴﺔ ﺃﺤﺩ ﺃﺴﺎﻟﻴﺏ ﺍﻟﺫﻜﺎﺀ ﺍﻻﺼﻁﻨﺎﻋﻲ‬
‫ﻭﻫﻲ ﻤﻥ ﺍﻷﺴﺎﻟﻴﺏ ﺍﻟﺤﺩﻴﺜﺔ ‪ ،‬ﺇﺫ ﺒﺭﺯﺕ ﺃﻫﻤﻴﺔ ﺍﺴﺘﺨﺩﺍﻡ ﻫﺫﺍ ﺍﻷﺴﻠﻭﺏ ﻓﻲ ﺤـل ﻤـﺴﺎﺌل ﻤﻌﻘـﺩﺓ‬
‫)ﻜﺒﻴﺭﺓ ﺍﻟﺤﺠﻡ ﺘﻤﺘﻠﻙ ﻜﻤﺎ ﻫﺎﺌﻼ ﻤﻥ ﺍﻟﺤﻠﻭل ﺍﻟﺒﺩﻴﻠﺔ( ﺨﻼل ﺯﻤﻥ ﻤﻨﺎﺴﺏ]‪.[2][3‬‬
‫ﻭﺘﻌﺩ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﺍﻟﺠﻴﻨﻴﺔ ﺍﻓﻀل ﻤﻥ ﺘﻘﻨﻴﺎﺕ ﺍﻟﺫﻜﺎﺀ ﺍﻻﺼﻁﻨﺎﻋﻲ ﺍﻟﺘﻘﻠﻴﺩﻴﺔ ﻟﻜﻭﻥ ﺍﻻﺨﻴـﺭﺓ‬
‫ﻤﻌﻘﺩﺓ ﻭﺘﻤﻴل ﺍﻟﻰ ﺍﻟﻭﻗﻭﻉ ﻓﻲ ﺍﻟﻤﻨﺎﻁﻕ ﺫﺍﺕ ﺍﻟﻨﻬﺎﻴﺎﺕ ﺍﻟﻤﺤﻠﻴـﺔ ﺍﻟـﺼﻐﺭﻯ )‪(Local minima‬‬
‫ﺍﻀﺎﻓﺔ ﺍﻟﻰ ﺍﻥ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﺍﻟﺠﻴﻨﻴﺔ ﻻﺘﻔﺸل ﻓﻲ ﺍﻴﺠﺎﺩ ﺍﻟﺤل ﺒﻭﺠﻭﺩ ﺘﻐﻴﻴﺭ ﻓﻲ ﺍﻻﺩﺨـﺎﻻﺕ ﻜـﺫﻟﻙ‬
‫ﻫﻲ ﺍﻓﻀل ﻓﻲ ﻋﻤﻠﻴﺔ ﺍﻟﺒﺤﺙ ﻀﻤﻥ ﻤﺠـﺎل ﺍﻟﻔـﻀﺎﺀ ﺍﻟﻭﺍﺴـﻊ )‪ (Complex space‬ﺍﻟﻤﺘﻌـﺩﺩ‬
‫ﺍﻻﺸﻜﺎل ﻭﻴﻜﻭﻥ ﺍﻟﺤل ﺍﻟﻨﺎﺘﺞ ﻤﻥ ﺘﻁﺒﻴﻕ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﺍﻟﺠﻴﻨﻴﺔ ﻓﻲ ﺍﻏﻠﺏ ﺍﻷﺤﻴﺎﻥ ﺤﻼ ﻗﺭﻴﺒﺎ ﺇﻟـﻰ‬
‫ﺍﻟﻤﺜﺎﻟﻲ )‪ ، (near optimal solution‬ﻭﻴﻭﻓﺭ ﻫﺫﺍ ﺍﻷﺴﻠﻭﺏ ﻋﻨﺩ ﺘﻁﺒﻴﻘﻪ ﺒﺤﺜﺎ ﺫﻜﻴـﺎ ﺒـﻴﻥ‬
‫ﻋﺩﺩ ﻫﺎﺌل ﻤﻥ ﺍﻟﺨﻁﻁ ﺍﻟﺒﺩﻴﻠﺔ]‪. [4‬‬

‫‪ .2‬ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﺍﻟﺠﻴﻨﻴﺔ‬
‫ﺘﻌﺭﻑ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﺍﻟﺠﻴﻨﻴﺔ ﺒﺄﻨﻬﺎ ﺨﻭﺍﺭﺯﻤﻴﺔ ﺫﻜﻴﺔ ﻴﻤﻜﻥ ﺍﺴﺘﺨﺩﺍﻤﻬﺎ ﻹﻴﺠﺎﺩ ﺤل ﺍﻟﻤـﺴﺎﺌل‬
‫ﺍﻟﻤﻌﻘﺩﺓ ﻭﺘﺤﺴﻴﻨﻬﺎ ‪ ،‬ﻭﺍﻟﺘﻲ ﺘﺩﺨل ﻓﻲ ﺍﻟﻌﺩﻴﺩ ﻤﻥ ﺍﻟﻤﺠﺎﻻﺕ ﻭﺘﻌﺩ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﺍﻟﺠﻴﻨﻴﺔ ﻤﻥ ﻁﺭﺍﺌـﻕ‬
‫ﺍﻟﺒﺤﺙ ﺍﻟﻜﻔﻭﺀﺓ ﺍﻟﻤﻌﺘﻤﺩﺓ ﻋﻠﻰ ﻤﺒﺎﺩﺉ ﺍﻻﺨﺘﻴﺎﺭ ﺍﻟﻁﺒﻴﻌﻲ ﻭﻋﻠﻡ ﺍﻟﻭﺭﺍﺜﺔ ‪ ،‬ﺍﺒﺘﻜﺭﻫـﺎ ﺍﻟﻌـﺎﻟﻡ ﺠـﻭﻥ‬
‫ﻫﻭﻻﻨﺩ )‪ (John Holland‬ﻋﺎﻡ ‪ 1975‬ﻓﻲ ﺠﺎﻤﻌﺔ ﻤﻴﺸﻴﻜﺎﻥ )‪، (University of Michigan‬‬
‫ﻭﻗﺩ ﻨﺸﺭ ﺒﺤﻭﺜﹰﺎ ﻋﺩﻴﺩﺓ ﻓﻲ ﻫﺫﺍ ﺍﻟﻤﺠﺎل ‪ ،‬ﻭﻜﺎﻥ ﺍﻟﻬـﺩﻑ ﺍﻷﺴـﺎﺱ ﻤﻨﻬـﺎ ﺒﻨـﺎﺀ ﺍﻟﻌﺩﻴـﺩ ﻤـﻥ‬
‫ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺎﺕ ﻭﺍﻟﺒﺭﻤﺠﻴﺎﺕ ﻭﺍﻷﻨﻅﻤﺔ ﻭﺘﺤﺴﻴﻨﻬﺎ ﺒﺎﺴﺘﺨﺩﺍﻡ ﻫﺫﻩ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ]‪.[8][5‬‬
‫ﺘﻁﺒﻕ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﺍﻟﺠﻴﻨﻴﺔ ﺒﻨﺠﺎﺡ ﻹﻴﺠﺎﺩ ﺍﻟﺤل ﺍﻟﻤﻘﺒﻭل )ﺍﻟﻘﺭﻴﺏ ﺇﻟﻰ ﺍﻟﻤﺜﺎﻟﻲ( ﻓﻲ ﺍﻟﻤﺴﺎﺌل‬
‫ﺍﻟﻤﺘﻌﻠﻘﺔ ﺒﺎﻟﻌﻠﻭﻡ ﻭﻤﻨﻬﺎ ﺍﻟﻌﻠﻭﻡ ﺍﻟﻁﺒﻴﺔ ﻭﺍﻟﻬﻨﺩﺴﻴﺔ ‪ ،‬ﺫﻟﻙ ﺃﻨﻬﺎ ﺍﺨﺘﺼﺭﺕ ﻜﺜﻴﺭﹰﺍ ﻤﻥ ﺍﻟﺯﻤﻥ ﻭﺍﻟﺠﻬـﺩ‬
‫ﺍﻟﻤﻁﻠﻭﺒﻴﻥ ﻟﺩﻯ ﻤﺼﻤﻤﻲ ﺍﻷﻨﻅﻤﺔ ﻭﺍﻟﺒﺭﺍﻤﺞ ‪ ،‬ﻭﺫﻟﻙ ﻤﻥ ﺨﻼل ﺇﻴﺠﺎﺩﻫﺎ ﺨﻭﺍﺭﺯﻤﻴﺔ ﻋﺎﻤﺔ ﻴﻌﺘﻤـﺩ‬
‫ﻋﻠﻴﻬﺎ ﻓﻲ ﺤل ﻤﺨﺘﻠﻑ ﺃﻨﻭﺍﻉ ﺍﻟﻤﺴﺎﺌل ‪ ،‬ﺒﺩﻻ ﻤﻥ ﺒﻨﺎﺀ ﺨﻭﺍﺭﺯﻤﻴﺔ ﺨﺎﺼﺔ ﻟﻜل ﻤﺴﺎﻟﺔ ‪ ،‬ﻤﻊ ﻤﺭﺍﻋﺎﺓ‬
‫ﺍﻟﺘﻐﻴﻴﺭﺍﺕ ﺍﻟﻼﺯﻤﺔ ﺍﻟﺘﻲ ﺘﺘﻨﺎﺴﺏ ﻤﻊ ﺨﺼﻭﺼﻴﺔ ﻜل ﻤﺴﺎﻟﺔ ﻤﻥ ﺤﻴﺙ ﺍﻟﺤﺠـﻡ ﻭﻨـﻭﻉ ﺍﻟﺒﻴﺎﻨـﺎﺕ‬
‫ﺍﻟﻤﺴﺘﺨﺩﻤﺔ ﻭﻁﺒﻴﻌﺔ ﺩﺍﻟﺔ ﺍﻟﻬﺩﻑ ﻭﺍﻟﻘﻴﻭﺩ ﻟﻜل ﻤﺴﺎﻟﺔ]‪. [1][7‬‬
‫‪ .3‬ﻋﻨﺎﺼﺭ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﺍﻟﺠﻴﻨﻴﺔ‬
‫ﺘﺘﺄﻟﻑ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﺍﻟﺠﻴﻨﻴﺔ ﻤﻥ ﻋﺩﺩ ﻤﻥ ﺍﻟﻌﻨﺎﺼﺭ ﻭﻓﻴﻤﺎ ﻴﺄﺘﻲ ﺘﻭﻀﻴﺢ ﻟﻬﺫﻩ ﺍﻟﻌﻨﺎﺼﺭ‪:‬‬

‫‪85‬‬
‫‪85‬‬
 ‫ﻤﻬﺎ ﻋﺒﺩ ﺍﻹﻟﻪ ﻤﺤﻤﺩ ﺍﻟﺒﺩﺭﺍﻨﻲ‬

‫‪ 1. 3‬ﺍﻟﻤﺠﺘﻤﻊ )‪(population‬‬
‫ﻴﺘﻜﻭﻥ ﺍﻟﻤﺠﺘﻤﻊ ﻤﻥ ﻋﺩﺩ ﻤﻥ ﺍﻷﻓﺭﺍﺩ ‪ ،‬ﺍﺫ ﻴﺤﺩﺩ ﻤﺼﻤﻡ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﻋﺩﺩ ﺍﻻﻓـﺭﺍﺩ‪ ،‬ﻓـﺈﺫﺍ‬
‫ﻜﺎﻥ ﺤﺠﻡ ﺍﻟﻤﺠﺘﻤﻊ ﻜﺒﻴﺭﺍ ﺠﺩﺍ ﺍﺤﻴﺎﻨﹰﺎ ﻓﺎﻨﻪ ﻗﺩ ﻻﻴﻌﻁﻲ ﺃﺩﺍﺀ ﺠﻴﺩﺍ ﻟﻠﺨﻭﺍﺭﺯﻤﻴﺔ ﻭﻜـﺫﻟﻙ ﺍﻟﺤـﺎل ﺇﺫﺍ‬
‫ﻜﺎﻥ ﺤﺠﻡ ﺍﻟﻤﺠﺘﻤﻊ ﺼﻐﻴﺭﺍ ﺠﺩﺍ ‪،‬ﺇﻥ ﻜل ﻓﺭﺩ )ﻜﺭﻭﻤﻭﺴﻭﻡ( ﻴﺘﻜﻭﻥ ﻤﻥ ﻋﺩﺩ ﻤﻥ ﺍﻟﻘـﻴﻡ ﻭﻴﻜـﻭﻥ‬
‫ﻋﺩﺩﹰﺍ )ﻁﻭل ﺍﻟﻜﺭﻭﻤﻭﺴﻭﻡ( ﻤﺤﺩﺩﹰﺍ ﺤﺴﺏ ﺍﻟﻤﺴﺎﻟﺔ ﻭﺘﻜﻭﻥ ﻗﻴﻤﺔ ﺍﻟﺒﺩﺍﻴﺔ ﻋﺸﻭﺍﺌﻴﺔ ﻀﻤﻥ ﻤﺤـﺩﺩﺍﺕ‬
‫ﺍﻟﻤﻌﺭﻓﺔ ﻤﻥ ﺍﻟﻤﺴﺎﻟﺔ ﺍﻟﻤﺭﺍﺩ ﺤﻠﻬﺎ ﻭﺍﻟﺸﻜل ﺁﻻﺘﻲ ﻴﻭﻀﺢ ﻫﻴﺌﺔ ﺍﻟﻜﺭﻭﻤﻭﺴﻭﻡ ﺍﺫ ﺇﻥ]‪: [7][9‬‬
‫‪ : N‬ﻫﻲ ﻋﺩﺩ ﺍﻷﻓﺭﺍﺩ ﻓﻲ ﺍﻟﻤﺠﺘﻤﻊ‪.‬‬
‫‪ : i‬ﻫﻭ ﺍﻟﻜﺭﻭﻤﻭﺴﻭﻡ )ﺍﻟﻔﺭﺩ( ﺍﻟﻤﺭﺍﺩ ﺘﻤﺜﻴﻠﻪ ‪ ،‬ﺍﺫ ﺇﻥ ‪. 1<=i<=N‬‬
‫‪ : j‬ﻫﻭ ﻁﻭل ﺍﻟﻜﺭﻭﻤﻭﺴﻭﻡ ‪ ،‬ﺍﺫ ﺇﻥ ‪. 1<=j<=I‬‬
‫‪Chromosomei = gene1 , gene2 , gene3 , ….. genej‬‬
‫‪ 2 .3‬ﺍﻟﺘﻘﻴﻴﻡ )‪(Evaluation‬‬
‫ﻴﻘﻴﻡ ﺍﻟﻜﺭﻭﻤﻭﺴﻭﻡ ﻓﻲ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﺍﻟﺠﻴﻨﻴﺔ ﺒﺩﺍﻟﺔ ﺍﻟﻠﻴﺎﻗﺔ ﻟﺘﺤﺩﻴﺩ ﻨﻭﻋﻴﺔ ﺍﻟﺤل ﻭﺘﻌﺭﻑ ﻫـﺫﻩ‬
‫ﺍﻟﺩﺍﻟﺔ ﺒﻬﺩﻑ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﺍﻟﺤﺎﻟﻴﺔ ﻟﻠﻤﺴﺎﻟﺔ ‪ ،‬ﻭﻟﺫﻟﻙ ﺘﻭﺠﺩ ﻟﻜل ﻓﺭﺩ ﺃﻭ ﻜﺭﻭﻤﻭﺴﻭﻡ ﺴـﻭﺍﺀ ﻜـﺎﻥ‬
‫ﻤﻭﻟﻭﺩﺍ ﻓﻲ ﺍﻟﻤﺠﺘﻤﻊ ﺍﻻﺒﺘﺩﺍﺌﻲ ﺃﻭ ﻤﻥ ﺨﻼل ﺍﻷﺠﻴﺎل ﺍﻟﻼﺤﻘﺔ ﻗﻴﻤﺔ ﺼﻼﺤﻴﺔ ﻭﻫﻲ ﻗﻴﻤﺔ ﺩﺍﻟﺔ ﺍﻟﻠﻴﺎﻗﺔ‬
‫ﻭﺘﺭﺘﺒﻁ ﻫﺫﻩ ﺍﻟﻘﻴﻤﺔ ﺒﻘﻴﻤﺔ ﺩﺍﻟﺔ ﺍﻟﻬـﺩﻑ ‪ ،‬ﻓـﺈﺫﺍ ﻜﺎﻨـﺕ ﻗﻴﻤـﺔ ﺍﻟـﺼﻼﺤﻴﺔ ﻟﻤـﺴﺎﺌل ﺍﻟﺘﻌﻅـﻴﻡ‬
‫)‪ (maximization problems‬ﻓﺈﻨﻬﺎ ﻤﺴﺎﻭﻴﺔ ﻟﻘﻴﻤﺔ ﺩﺍﻟﺔ ﺍﻟﻬﺩﻑ ‪ ،‬ﺃﻤﺎ ﻓﻲ ﻤـﺴﺎﺌل ﺍﻟﺘـﺼﻐﻴﺭ‬
‫)‪ (Minimization problems‬ﻓﺎﻥ ﺍﻟﻬﺩﻑ ﻫﻭ ﺇﻴﺠﺎﺩ ﺍﻟﺤل ﺫﻱ ﺍﻟﻘﻴﻤﺔ ﺍﻟﺩﻨﻴﺎ ﻟﺩﺍﻟﺔ ﺍﻟﻬﺩﻑ ﻭﺒﺫﻟﻙ‬
‫ﺘﺤﺼل ﺍﻟﺤﻠﻭل ﺫﻭﺍﺕ ﺍﻟﻘﻴﻡ ﺍﻟﺼﻐﺭﻯ ﻟﺩﻭﺍل ﺍﻟﻬﺩﻑ ﻋﻠﻰ ﻗﻴﻤﺔ ﺍﻜﺒﺭ ﻟﻠﺼﻼﺤﻴﺔ ]‪.[4‬‬

‫‪ 3. 3‬ﺍﻟﺘﺸﻔﻴﺭ )‪(Encoding‬‬
‫ﻴﻘﺼﺩ ﺒﺎﻟﺘﺸﻔﻴﺭ ﻋﻤﻠﻴﺔ ﺇﻴﺠﺎﺩ ﺘﻤﺜﻴل ﻤﻨﺎﺴﺏ ﻟﻠﺤل ‪ ،‬ﺍﺫ ﻴﻌﺩ ﺍﻟﺘﺸﻔﻴﺭ ﻋﺎﻤﻼ ﺃﺴﺎﺴﻴﺎ ﻟﻨﺠـﺎﺡ‬
‫ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﺍﻟﺠﻴﻨﻴﺔ ‪ ،‬ﻭﻴﻌﺘﻤﺩ ﻋﻠﻰ ﺍﻟﻤﺴﺎﻟﺔ ﺍﻟﻤﺭﺍﺩ ﺤﻠﻬﺎ‪ ،‬ﻓﻬﻨﺎﻟﻙ ﻋﺩﺓ ﻁﺭﻕ ﻟﻌﻤﻠﻴﺔ ﺍﻟﺘﺸﻔﻴﺭ ﻭﻫﺫﺍ‬
‫ﻴﻌﺘﻤﺩ ﻋﻠﻰ ﺍﻟﻤﺴﺄﻟﺔ ﺍﻟﻤﺭﺍﺩ ﺤﻠﻬﺎ ﻻﻥ ﺍﻟﻜﺭﻭﻤﻭﺴﻭﻡ ﻴﺠﺏ ﺍﻥ ﻴﺤﺘﻭﻱ ﻋﻠﻰ ﻤﻌﻠﻭﻤﺎﺕ ﺤﻭل ﺍﻟﺤـل‬
‫ﻟﻬﺫﺍ ﻴﺠﺏ ﺍﻥ ﻴﺸﻔﺭ ﺒﻁﺭﻴﻘﺔ ﻤﻨﺎﺴﺒﺔ ﻭﻤﻥ ﻫﺫﻩ ﺍﻟﻁﺭﻕ ]‪: [10‬‬
‫‪ 1 .3. 3‬ﺍﻟﺘﺸﻔﻴﺭ ﺍﻟﺜﻨﺎﺌﻲ )ﺅ(‬
‫ﻫﺫﺍ ﺍﻟﻨﻭﻉ ﻤﻥ ﺍﻟﺘﺸﻔﻴﺭ ﻫﻭ ﺍﻷﻜﺜﺭ ﺸﻴﻭﻋﺎ ﻭﺫﻟﻙ ﻻﻨﻪ ﺍﻟﺘﺸﻔﻴﺭ ﺍﻷﻭل ﺍﻟﺫﻱ ﺍﺴـﺘﺨﺩﻡ ﻤـﻊ‬
‫ﻼ ﻋﻥ ﺒﺴﺎﻁﺘﻪ ﺍﻟﻨﺴﺒﻴﺔ ‪ ،‬ﻓﻔﻲ ﻫﺫﺍ ﺍﻟﻨﻭﻉ ﻴﺘﻜﻭﻥ ﺍﻟﻜﺭﻭﻤﻭﺴﻭﻡ ﻤﻥ ﺴﻠﺴﻠﺔ‬
‫ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﺍﻟﺠﻴﻨﻴﺔ ﻓﻀ ﹰ‬
‫ﺃﺭﻗﺎﻡ ‪ 1‬ﻭ‪ 0‬ﻭﻴﻭﻀﺢ ﺍﻟﺸﻜل )‪ (1‬ﻜﺭﻭﻤﻭﺴﻭﻤﺎ ﻤﺸﻔﺭﺍ ﺒﻬﺫﺍ ﺍﻟﻨﻭﻉ ﻤﻥ ﺍﻟﺘﺸﻔﻴﺭ ]‪:[3][2‬‬
‫‪1‬‬ ‫‪0‬‬ ‫‪1‬‬ ‫‪0‬‬ ‫‪0‬‬ ‫‪……………… 1‬‬
‫ﺍﻟﺸﻜل )‪ (1‬ﺍﻟﺘﺸﻔﻴﺭ ﺍﻟﺜﻨﺎﺌﻲ‬

‫‪86‬‬
‫‪86‬‬
 ‫ﺍﺴﺘﺨﺩﺍﻡ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﺍﻟﺠﻴﻨﻴﺔ ﻓﻲ ﺘﻁﺎﺒﻕ ‪....‬‬

‫‪ 2 .3. 3‬ﺍﻟﺘﺸﻔﻴﺭ ﺃﻻﺴﺘﺒﺩﺍﻟﻲ )‪(Permutation Encoding‬‬

‫ﻓﻲ ﻫﺫﺍ ﺍﻟﻨﻭﻉ ﻤﻥ ﺍﻟﺘﺸﻔﻴﺭ ﻜل ﻤﻭﻗﻊ ﺒﺎﻟﻜﺭﻭﻤﻭﺴﻭﻡ ﻴﺤﻤل ﺭﻗﻤﹰﺎ ﺒﻴﻥ ‪.[10] 0,9‬‬

‫‪4‬‬ ‫‪5‬‬ ‫‪3‬‬ ‫‪6‬‬ ‫‪9‬‬ ‫‪7‬‬ ‫‪2‬‬ ‫‪1‬‬

‫ﺍﻟﺸﻜل )‪ (2‬ﺍﻟﺘﺸﻔﻴﺭ ﺍﻻﺴﺘﺒﺩﺍﻟﻲ‬

‫‪ 3 .3. 3‬ﺍﻟﺘﺸﻔﻴﺭ ﺍﻟﺤﻘﻴﻘﻲ )‪(Real value Encoding‬‬

‫ﻓﻲ ﻫﺫﺍ ﺍﻟﻨﻭﻉ ﻤﻥ ﺍﻟﺘﺸﻔﻴﺭ ﻜل ﻤﻭﻗﻊ ﺒﺎﻟﻜﺭﻭﻤﻭﺴﻭﻡ ﻴﻤﺜل ﺭﻗﻤﹰﺎ ﺤﻘﻴﻘﻴﹰﺎ ]‪.[3‬‬

‫‪1.29 5.68 2.73‬‬ ‫‪4.44‬‬

‫ﺍﻟﺸﻜل )‪ (3‬ﺍﻟﺘﺸﻔﻴﺭ ﺍﻟﺤﻘﻴﻘﻲ‬

‫‪ 4 .3. 3‬ﺍﻟﺘﺸﻔﻴﺭ ﺍﻟﺸﺠﺭﻱ )‪(Tree Encoding‬‬

‫ﺍﻟﻜﺭﻭﻤﻭﺴﻭﻡ ﻴﺘﻤﺜل ﺒﺸﻜل ﺸﺠﺭﺓ ﻤﺘﻜﻭﻨﺔ ﻤﻥ ﻋﺩﺩ ﻤﻥ ﺍﻟﻌﻘﺩ )‪ (nodes‬ﻭﺍﻟﺤﺎﻓﺎﺕ)‪(arcs‬‬
‫ﻭﻜل ﻋﻘﺩﺓ ﺘﺘﻤﺜل ﺒﻌﻤﻠﻴﺔ ﺭﻴﺎﻀﻴﺔ ﺍﻭ ﻤﺘﻐﻴﺭ ﺍﻭ ﻗﻴﻤﺔ ﺜﺎﺒﺘﺔ ]‪.[2‬‬

‫‪+‬‬

‫‪X‬‬ ‫‪-‬‬

‫‪y‬‬ ‫‪8‬‬

‫ﺍﻟﺸﻜل )‪ (4‬ﺍﻟﺘﺸﻔﻴﺭ ﺍﻟﺸﺠﺭﻱ‬

‫‪ 4. 3‬ﺍﻻﻨﺘﻘﺎﺀ )‪(Selection‬‬
‫ﻴﻘﺼﺩ ﺒﺎﻻﻨﺘﻘﺎﺀ ﻋﻤﻠﻴﺔ ﺍﺨﺘﻴﺎﺭ ﺍﻷﺒﺎﺀ ﻤﻥ ﺍﻟﻤﺠﺘﻤﻊ ﻷﺠل ﺍﻟﺘﺯﺍﻭﺝ ﻭﺇﻨﺘﺎﺝ ﺠﻴل ﺠﺩﻴﺩ‪ ،‬ﻓﺒﻌـﺩ‬
‫ﺍﺨﺘﻴﺎﺭ ﻁﺭﻴﻘﺔ ﺍﻟﺘﺸﻔﻴﺭ ﺍﻟﻤﻨﺎﺴﺒﺔ ﻴﺼﺒﺢ ﺍﻟﻘﺭﺍﺭ ﺍﻟﺫﻱ ﻴﺠﺏ ﺍﺘﺨﺎﺫﻩ ﻫﻭ ﻜﻴﻔﻴﺔ ﺇﺠﺭﺍﺀ ﺍﻻﻨﺘﻘﺎﺀ‪.‬‬
‫ﻭﻁﺒﻘﺎ ﻟﻨﻅﺭﻴﺔ ﺩﺍﺭﻭﻥ ﺍﻟﺘﻲ ﺘﻨﺹ ﻋﻠﻰ ﻤﺒﺩﺃ ﺍﻟﺒﻘـﺎﺀ ﻟﻸﺼـﻠﺢ‪ ،‬ﻴﺤـﺼل ﺍﻷﻓـﺭﺍﺩ ﺫﻭﻭ‬
‫ﺍﻟﺼﻼﺤﻴﺔ ﺍﻷﻋﻠﻰ ﻋﻤﻭﻤﹰﺎ ﻋﻠﻰ ﻓﺭﺼﺔ ﺍﻜﺒﺭ ﻟﻠﻌﻴﺵ ﻭﺍﻟﺘﺯﺍﻭﺝ ﻤـﻥ ﺍﻷﻓـﺭﺍﺩ ﺫﻭﻱ ﺍﻟـﺼﻼﺤﻴﺔ‬
‫ﺍﻟﻭﺍﻁﺌﺔ]‪ ,[3‬ﻭﻤﻥ ﻁﺭﺍﺌﻕ ﺍﻻﻨﺘﻘﺎﺀ]‪:[5‬‬

‫‪87‬‬
‫‪87‬‬
 ‫ﻤﻬﺎ ﻋﺒﺩ ﺍﻹﻟﻪ ﻤﺤﻤﺩ ﺍﻟﺒﺩﺭﺍﻨﻲ‬

‫‪ 1 .4. 3‬ﺍﻻﻨﺘﻘﺎﺀ ﺍﻟﻨﺴﺒﻲ )‪(Proportional selection‬‬

‫ﻁﺭﻴﻘﺔ ﺍﻻﻨﺘﻘﺎﺀ ﺍﻟﻨﺴﺒﻲ )‪ (Proportional selection‬ﺍﺫ ﻴﺘﻡ ﺍﻻﻨﺘﻘﺎﺀ ﺒﺎﻻﻋﺘﻤـﺎﺩ ﻋﻠـﻰ‬
‫ﻨﺴﺒﺔ ﻗﻴﻤﺔ ﺍﻟﻠﻴﺎﻗﺔ ﻟﻜل ﻓﺭﺩ ﻤﻘﺴﻭﻤﺎ ﻋﻠﻰ ﻤﺠﻤﻭﻉ ﻗﻴﻡ ﺩﻭﺍل ﺍﻟﻠﻴﺎﻗﺔ ﻷﻓﺭﺍﺩ ﺍﻟﻤﺠﺘﻤﻊ ﻭﺫﻟﻙ ﺒﺤـﺴﺏ‬
‫ﺍﻟﻤﻌﺎﺩﻟﺔ ﺍﻻﺘﻴﺔ]‪:[5‬‬
‫)) ‪f ( xi(t‬‬
‫= )‪prop( x‬‬
‫)) ‪∑ f ( xi(t‬‬
‫ﺍﺫ ‪ i‬ﻴﻤﺜل ﺭﻗﻡ ﺍﻟﻜﺭﻭﻤﻭﺴﻭﻡ ‪.‬‬
‫)‪ f(x‬ﺘﻤﺜل ﻤﺩﻯ ﺍﻟﻠﻴﺎﻗﺔ ﻟﻠﻜﺭﻭﻤﻭﺴﻭﻡ ‪.‬‬
‫‪ t‬ﻴﻤﺜل ﺭﻗﻡ ﺍﻟﺠﻴل ‪.‬‬
‫ﺜﻡ ﻴﺘﻡ ﺍﻻﺨﺘﻴﺎﺭ ﺒﺤﺴﺏ ﺍﻟﻤﺴﺎﻟﺔ ﺃﻱ ﺍﻨﺘﺨﺎﺏ ﺯﻭﺝ ﻤﻥ ﺍﻟﻤﻘﺎﻁﻊ ﺍﻟﺘﻲ ﺘﻤﺘﻠﻙ ﺍﺤﺴﻥ ﻗﻴﻡ ﻟﻠﻴﺎﻗﺔ‪.‬‬
‫‪ 2 .4. 3‬ﺍﻨﺘﻘﺎﺀ ﻋﺠﻠﺔ ﺍﻟﺭﻭﻟﻴﺕ )‪(Roulette wheel selection‬‬
‫ﻤﻘﺎﻁﻊ ﺍﻻﺒﺎﺀ ﺘﺨﺘﺎﺭ ﺤﺴﺏ ﻗﻴﻡ ﺍﻟﻠﻴﺎﻗﺔ ﺍﻟﻤﺤﺴﻭﺒﺔ )‪ (fitness‬ﻭﺍﻥ ﺍﻟﻤﻘﻁﻊ ﺍﻻﺤﺴﻥ ﻴﻜﻭﻥ‬
‫ﺫﺍ ﺤﻅ ﺍﻭﻓﺭ ﻟﻠﻤﺴﺎﻫﻤﺔ ﻓﻲ ﺘﻜﻭﻴﻥ ﺍﻟﺠﻴل ﺍﻟﻼﺤﻕ ‪ .‬ﺍﺫ ﻴﺘﻡ ﺘﻤﺜﻴل ﺍﺤﺘﻤﺎﻟﻴﺔ ﻤﺴﺎﻫﻤﺔ ﻜل ﻤﻘﻁﻊ ﻋﻠﻰ‬
‫ﻋﺠﻠﺔ ﺍﻟﺭﻭﻟﻴﺕ ﻟﻜﻲ ﻴﺸﻐل ﺤﻴﺯﺍ ﻋﻠﻰ ﺍﻟﻌﺠﻠﺔ ﻭﺘﻤﺜل ﻋﻤﻠﻴﺔ ﺍﻨﺘﺨﺎﺏ ﺍﻟﻤﻘﻁﻊ ﺒﺘـﺩﻭﻴﺭ ﺍﻟﻌﺠﻠـﺔ ‪،‬‬
‫ﻓﺎﻟﺫﻱ ﻴﺸﻐل ﺤﻴﺯﺍ ﺍﻜﺒﺭ ﻓﺎﺤﺘﻤﺎﻟﻴﺔ ﺍﻨﺘﺨﺎﺒﻪ ﺍﻋﻠﻰ ﻤﻥ ﻏﻴﺭﻩ ﻤﻥ ﺍﻟﻤﻘﺎﻁﻊ ‪.‬‬

‫ﺸﻜل )‪ (5‬ﺍﻨﺘﻘﺎﺀ ﻋﺠﻠﺔ ﺍﻟﺭﻭﻟﻴﺕ‬

‫‪ 3 .4. 3‬ﺍﻻﻨﺘﻘﺎﺀ ﺍﻟﺼﻔﻲ )‪(Rank selection‬‬

‫ﻻ ﺘﺭﺘﻴﺏ ﺍﻟﻜﺭﻭﻤﻭﺴﻭﻤﺎﺕ ﺤﺴﺏ ﺩﺭﺠﺔ ﺍﻟﻠﻴﺎﻗﺔ ﻟﻜل‬
‫ﻓﻲ ﻫﺫﻩ ﺍﻟﻁﺭﻴﻘﺔ ﻤﻥ ﺍﻻﺨﺘﻴﺎﺭ ﻴﺘﻡ ﺍﻭ ﹰ‬
‫ﻭﺍﺤﺩ ﺘﺭﺘﻴﺏ ﺘﺼﺎﻋﺩﻱ ﻓﻜل ﻜﺭﻭﻤﻭﺴﻭﻡ ‪ X‬ﻟﻪ ﺭﺘﺒﺔ ))‪ (Rank(x‬ﺃﻱ ﺍﻟﻤﻭﻗﻊ ﺒﻌﺩ ﺍﻟﺘﺭﺘﻴﺏ‬
‫ﻭﺒﻌﺩﻫﺎ ﻴﺘﻡ ﺤﺴﺎﺏ ﻨﺴﺒﺔ ﺍﻟﻬﺩﻑ ‪ (Target sampling rate) TSR‬ﻟﻜل ﻜﺭﻭﻤﻭﺴﻭﻡ ﻤﻥ‬
‫ﺍﻟﻤﻌﺎﺩﻟﺔ ﺍﻻﺘﻴﺔ‪:‬‬
‫‪Rank(x)-1‬‬
‫)‪TSR (x) =Min + (Max- Min‬‬
‫‪N-1‬‬

‫‪88‬‬
‫‪88‬‬
 ‫ﺍﺴﺘﺨﺩﺍﻡ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﺍﻟﺠﻴﻨﻴﺔ ﻓﻲ ﺘﻁﺎﺒﻕ ‪....‬‬

‫ﺤﻴﺙ ‪ N‬ﻫﻲ ﻋﺩﺩ ﺍﻟﻜﺭﻭﻤﻭﺴﻭﻤﺎﺕ ﺒﺎﻟﺠﻴل‪ ،‬ﻓﺎﻻﺴﻭﺀ ﻴﺄﺨﺫ ﺍﻟﻘﻴﻤﺔ ‪ 1‬ﻭﺍﻟﺜﺎﻨﻲ ﻴﺄﺨﺫ ‪ 2‬ﻤﻥ ﻗﻴﻤﺔ‬
‫‪ TSR‬ﺍﻟﻰ ﺃﻥ ﻨﺼل ﺍﻟﻰ ﺍﻷﺤﺴﻥ‪.‬‬

‫‪ 5. 3‬ﺍﻟﺘﻘﺎﻁﻊ )‪(Crossover‬‬
‫ﺍﻟﺘﻘﺎﻁﻊ ﻓﻲ ﻋﻤﻠﻴﺎﺕ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﺍﻟﺠﻴﻨﻴﺔ ﻫﻭ ﺩﻤﺞ ﻜﺭﻭﻤﻭﺴـﻭﻤﻴﻥ )ﻭﺍﻟـﺩﻴﻥ( ﻟﺘﻜـﻭﻴﻥ‬
‫ﻜﺭﻭﻤﻭﺴﻭﻡ ﺠﺩﻴﺩ )ﺍﺒﻥ( ‪ ,‬ﻭﺍﻟﻔﻜﺭﺓ ﻤﻥ ﻫﺫﻩ ﺍﻟﻌﻤﻠﻴﺔ ﻫﻲ ﺃﻥ ﺍﻟﻜﺭﻭﻤﻭﺴﻭﻡ ﺍﻟﻨﺎﺘﺞ ﻴﻜﻭﻥ ﺃﻓﻀل ﻤﻥ‬
‫ﺍﻷﺒﻭﻴﻥ ﺇﺫﺍ ﺍﺨﺫ ﺍﻟﺼﻔﺎﺕ ﺍﻟﺠﻴﺩﺓ ﻤﻥ ﻜﻠﻴﻬﻤﺎ ‪ ,‬ﻭﻤﻥ ﺼﻴﻎ ﺍﻟﺘﻘﺎﻁﻊ ﺍﻟﺘﻘـﺎﻁﻊ ﺫﻭ ﻨﻘﻁﺘـﻲ ﺍﻟﻘﻁـﻊ‬
‫)‪ (Two – point crossover‬ﻭﻓﻲ ﻫﺫﺍ ﺍﻟﻨﻭﻉ ﻨﺤﺘﺎﺝ ﺇﻟﻰ ﺍﺨﺘﻴﺎﺭ ﻨﻘﻁﺘﻲ ﺘﻘـﺎﻁﻊ ﻋـﺸﻭﺍﺌﻴﺘﻴﻥ‬
‫ﻀﻤﻥ ﺤﺩﻭﺩ )ﻁﻭل( ﺍﻟﻜﺭﻭﻤﻭﺴﻭﻡ ﺜﻡ ﻨﺒﺩل ﺍﻟﻜﺭﻭﻤﻭﺴﻭﻤﺎﺕ ﻟﻸﺒﻭﻴﻥ ﺒﻴﻥ ﻫﺎﺘﻴﻥ ﺍﻟﻨﻘﻁﺘﻴﻥ ﻹﻨﺘـﺎﺝ‬
‫ﻁﻔﻠﻴﻥ ﺠﺩﻴﺩﻴﻥ ‪ .‬ﻜﻤﺎ ﻓﻲ ﺍﻟﻤﺜﺎل ﺍﻵﺘﻲ ﺍﻟﺫﻱ ﺍﺴﺘﺨﺩﻡ ﻓﻴﻪ ﺍﻟﺘﺸﻔﻴﺭ ﺍﻟﺜﻨﺎﺌﻲ]‪: [1‬‬
‫‪Parent 1 : 110 / 010 / 10‬‬
‫‪Parent 2: 001/ 001 / 11‬‬
‫ﺤﻴﺙ ﺇﻥ ﺍﻹﺸﺎﺭﺓ ‪ /‬ﺘﻤﺜل ﻤﻭﻗﻊ ﻨﻘﻁﺘﻲ ﺍﻟﺘﻘﺎﻁﻊ ﻓﺒﻌﺩ ﺘﺒﺩﻴل ﻤﺎ ﺒﻴﻥ ﺍﻟﻨﻘﻁﺘﻴﻥ ﻴﻜﻭﻥ ﻟﺩﻴﻨﺎ ﻁﻔﻼﻥ‬
‫ﺠﺩﻴﺩﺍﻥ ‪:‬‬
‫‪Child 1 : 110/001/10‬‬
‫‪Child 2 : 001/010/11‬‬
‫‪ 6. 3‬ﺍﻟﻁﻔﺭﺓ )‪(Mutation‬‬
‫ﺍﻟﻁﻔﺭﺓ ﻫﻲ ﻋﻤﻠﻴﺔ ﺘﻐﻴﻴﺭ ﻓﻲ ﻗﻴﻡ ﺍﻟﺠﻴﻥ ﻟﻠﻜﺭﻭﻤﻭﺴﻭﻡ ﺍﻟﻨﺎﺘﺞ ﻤﻥ ﻋﻤﻠﻴﺔ ﺍﻟﺘﻘﺎﻁﻊ ﻭﻴﺘﻡ ﻫـﺫﺍ‬
‫ﺍﻟﺘﻐﻴﻴﺭ ﺒﺎﻻﻋﺘﻤﺎﺩ ﻋﻠﻰ ﻨﺴﺒﺔ ﺍﻟﻁﻔﺭﺓ ﺍﻟﺘﻲ ﻋﺎﺩﺓ ﻤﺎ ﺘﻜﻭﻥ ﻗﻴﻤﺘﻬﺎ ﺍﻟﻌﺩﺩﻴﺔ ﺼﻐﻴﺭﺓ ﺠﺩﺍ )‪ 0,01‬ﺍﻭ‬
‫ﻻ ﺠﺩﻴﺩﺓ )ﻟﻡ ﻴﺴﺒﻕ ﺘﻜﻭﻴﻨﻬﺎ ﻓﻲ ﺍﻷﺠﻴﺎل ﺍﻟﺴﺎﺒﻘﺔ( ﺘﻀﺎﻑ‬
‫ﺍﻗل( ﻭﻓﻲ ﺤﺎﻟﺔ ﺤﺼﻭل ﺍﻟﻁﻔﺭﺓ ﻓﺎﻥ ﺤﻠﻭ ﹰ‬
‫ﺇﻟﻰ ﺍﻟﻤﺠﺘﻤﻊ ﺍﻟﺠﺩﻴﺩ )ﻗﺩ ﻴﻜﻭﻨﻭﻥ ﺍﺤﺴﻥ ﺃﻭ ﺍﺴﻭﺃ ﻤﻥ ﺍﻷﻓﺭﺍﺩ ﺍﻟﺴﺎﺒﻘﻴﻥ( ﻭﺘﻬﺩﻑ ﻫﺫﻩ ﺍﻟﻌﻤﻠﻴﺔ ﺇﻟﻰ‬
‫ﺘﻭﺴﻴﻊ ﻤﺠﺎل ﺍﻟﺤﻠﻭل ﺍﻟﻤﻤﻜﻨﺔ ﻭﺍﻟﻤﺘﻤﺜل ﺒﺘﻜﻭﻴﻥ ﺍﻜﺒﺭ ﻋﺩﺩ ﻤﻥ ﺍﻷﻓﺭﺍﺩ ﺍﻟﻤﺨﺘﻠﻔﻴﻥ ﻀﻤﻥ ﺍﻟﻤﺠﺘﻤـﻊ‬
‫ﺫﻟﻙ ﺃﻥ ﺍﻷﻓﺭﺍﺩ ﻜﻠﻤﺎ ﻜﺎﻨﻭﺍ ﻤﺨﺘﻠﻔﻴﻥ ﻓﻲ ﺍﻟﻤﺠﺘﻤﻊ ‪ ،‬ﻓﺎﻥ ﻫﺫﺍ ﻴﻭﻓﺭ ﺤﻠﻭﻻ ﺇﻀـﺎﻓﻴﺔ ﻤﻤـﺎ ﻴـﺅﺜﺭ‬
‫ﺍﻴﺠﺎﺒﻴﺎ ﻓﻲ ﺴﺭﻋﺔ ﺍﻗﺘﺭﺍﺏ ﺍﻟﺤل ‪ ،‬ﻭﻤﻥ ﺼﻴﻎ ﺍﻟﻁﻔﺭﺓ ﺍﻟﻭﺭﺍﺜﻴﺔ]‪: [8‬‬
‫‪ .1‬ﺒﺕ )‪ : (Bit‬ﺘﺴﺘﻌﻤل ﻤﻊ ﺍﻟﺘﺸﻔﻴﺭ ﺍﻟﺜﻨﺎﺌﻲ )‪ ، (Binary Encoding‬ﺤﻴﺙ ﻨﻘﻭﻡ ﺒﺎﺴـﺘﺒﺩﺍل ﻗﻴﻤـﺔ‬
‫)‪ (Gene‬ﻤﻥ ﺍﻟﻜﺭﻭﻤﻭﺴﻭﻡ ﺤﻴﺙ ﻴﺨﺘﺎﺭ ﻤﻭﻗﻌﺎ ﻋﺸﻭﺍﺌﻴﺎ ﻀﻤﻥ ﺤﺩﻭﺩ ﺍﻟﻜﺭﻭﻤﻭﺴﻭﻡ ﻓـﺎﺫﺍ ﻜﺎﻨـﺕ‬
‫ﺍﻟﻘﻴﻤﺔ ﻫﻲ ‪ 0‬ﺘﺴﺘﺒﺩل ﺒـ ‪ 1‬ﻭﺍﺫﺍ ﻜﺎﻨﺕ ‪ 1‬ﺘﺴﺘﺒﺩل ﺒـ ‪. 0‬‬
‫‪ .2‬ﺍﻟﺤﺩﻭﺩ )‪ : (Boundary‬ﺘﺴﺘﻌﻤل ﻤﻊ ﺍﻟﺘﺸﻔﻴﺭ ﺍﻟﺤﻘﻴﻘﻲ ﺍﻭ ﺍﻟﺼﺤﻴﺢ)‪(Real value Encoding‬‬
‫‪ ،‬ﺤﻴﺙ ﻨﻘﻭﻡ ﺒﺎﺴﺘﺒﺩﺍل ﻗﻴﻤﺔ )‪ (Gene‬ﻤﻥ ﺍﻟﻜﺭﻭﻤﻭﺴﻭﻡ ﻴﺨﺘﺎﺭ ﻤﻭﻗﻌﺎ ﻋـﺸﻭﺍﺌﻴﺎ ﻀـﻤﻥ ﺤـﺩﻭﺩ‬
‫ﺍﻟﻜﺭﻭﻤﻭﺴﻭﻡ ﻭﺘﺴﺘﺒﺩل ﺍﻟﻘﻴﻤﺔ ﺒﻘﻴﻤﺔ ﻋﺸﻭﺍﺌﻴﺔ ﻀﻤﻥ ﺤﺩﻭﺩ ﺍﻟﻘﻴﻡ ﺍﻟﻤﺘﺎﺤﺔ ﻟﻠﻜﺭﻭﻤﻭﺴﻭﻡ‪.‬‬

‫‪89‬‬
‫‪89‬‬
 ‫ﻤﻬﺎ ﻋﺒﺩ ﺍﻹﻟﻪ ﻤﺤﻤﺩ ﺍﻟﺒﺩﺭﺍﻨﻲ‬

‫‪ .3‬ﺍﻟﻤﻨﺘﻅﻡ )‪ : (Uniform‬ﺘﺴﺘﻌﻤل ﻤﻊ ﺍﻟﺘﺸﻔﻴﺭ ﺍﻟﺤﻘﻴﻘﻲ ﺍﻭ ﺍﻟﺼﺤﻴﺢ ‪ ،‬ﺤﻴﺙ ﻨﻘﻭﻡ ﺒﺎﺴـﺘﺒﺩﺍل ﻗﻴﻤـﺔ‬

‫)‪ (Gene‬ﻤﻥ ﺍﻟﻜﺭﻭﻤﻭﺴﻭﻡ ﻴﺨﺘﺎﺭ ﻤﻭﻗﻌﺎ ﻋﺸﻭﺍﺌﻴﺎ ﻀﻤﻥ ﺤﺩﻭﺩ ﺍﻟﻜﺭﻭﻤﻭﺴﻭﻡ ﻭﻨـﺴﺘﺒﺩل ﺍﻟﻘﻴﻤـﺔ‬
‫ﺒﻘﻴﻤﺔ ﻋﺸﻭﺍﺌﻴﺔ ﻭﻟﻜﻥ ﻀﻤﻥ ﺤﺩﻭﺩ ﻴﻌﺭﻓﻬﺎ ﺍﻟﻤﺴﺘﻌﻤل ‪.‬‬
‫‪ .4‬ﻤﻘﺎﻴﻴﺱ ﺍﻟﺘﻭﻗﻑ ﻓﻲ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﺍﻟﺠﻴﻨﻴﺔ‬
‫ﺇﻥ ﻤﻘﻴﺎﺱ ﺍﻟﺘﻭﻗﻑ ﻓﻲ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﺍﻟﺠﻴﻨﻴﺔ ﻴﺤﺩﺩ ﻤﺎ ﺍﺫﺍ ﻜﺎﻨﺕ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﺍﻟﺠﻴﻨﻴﺔ ﺴﺘﺴﺘﻤﺭ‬
‫ﻓﻲ ﺍﻟﺒﺤﺙ ﺃﻡ ﺴﺘﺘﻭﻗﻑ؟ ﻭﻴﻔﺤﺹ ﻤﻘﻴﺎﺱ ﺍﻟﺘﻭﻗﻑ ﻟﻠﺨﻭﺍﺭﺯﻤﻴﺔ ﺒﻌﺩ ﻜل ﺠﻴل ﻟﻨﺭﻯ ﺍﺫﺍ ﻤﺎ ﻜﺎﻨـﺕ‬
‫ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﺴﺘﺘﻭﻗﻑ ﺃﻡ ﻻ ؟ ﻭﻫﻨﺎﻙ ﻋﺩﺓ ﻤﻘﺎﻴﻴﺱ ﻟﺘﻭﻗﻑ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﺍﻟﺠﻴﻨﻴﺔ ﺘﺨﺘﻠـﻑ ﺒﺤـﺴﺏ‬
‫ﺍﻟﻤﺴﺄﻟﺔ ﺍﻟﻤﺭﺍﺩ ﺤﻠﻬﺎ ﻭﻤﻥ ﺃﻫﻡ ﺍﻟﻤﻌﺎﻴﻴﺭ ﺍﻟﻤﺴﺘﺨﺩﻤﺔ ﻗﻴﻤﺔ ﺩﺍﻟﺔ ﺍﻟﻠﻴـﺎﻗﺔ ﺍﺫ ﺘﺘﻭﻗـﻑ ﺍﻟﺨﻭﺍﺭﺯﻤﻴـﺔ‬
‫ﻋﻨﺩ ﺍﺤﺴﻥ ﻗﻴﻤﺔ ﻟﺩﺍﻟﺔ ﺍﻟﻠﻴﺎﻗﺔ ﻓﻲ ﺍﻟﻤﺠﺘﻤﻊ ﺍﻟﺤﺎﻟﻲ ﻋﻨﺩﻤﺎ ﺘﻜﻭﻥ ﺍﻗل ﻤﻥ ﻗﻴﻤﺔ ﺍﻟﻠﻴﺎﻗﺔ ﺍﻟﺘﻲ ﺤـﺩﺩﺕ‬
‫)ﻓﻲ ﺤﺎﻟﺔ ‪ (Minimize‬ﺃﻭ ﺘﺘﻭﻗﻑ ﻋﻨﺩﻤﺎ ﺘﻜﻭﻥ ﺃﺤﺴﻥ ﻗﻴﻤﺔ ﻟﺩﺍﻟﺔ ﺍﻟﻠﻴﺎﻗﺔ ﺍﻜﺒﺭ ﻤﻥ ﻗﻴﻤﺔ ﺍﻟﻠﻴـﺎﻗﺔ‬
‫ﺍﻟﺘﻲ ﺤﺩﺩﺕ )ﻓﻲ ﺤـﺎﻟﺔ ‪. [2] [4](Maximize‬‬

‫‪ .5‬ﺘﻤﻴﻴﺯ ﺍﻷﻨﻤﺎﻁ‬
‫ﻨﺘﻴﺠﺔ ﻟﻠﻀﻐﻁ ﺍﻟﺫﻱ ﻭﻟﺩﺘﻪ ﺍﻟﺘﻘﺎﻨﺎﺕ ﺍﻟﺤﺩﻴﺜﺔ ﺩﻋﺕ ﺍﻟﺤﺎﺠﺔ ﺇﻟﻰ ﻀﺭﻭﺭﺓ ﺒﻨﺎﺀ ﺍﻻﺕ ﺘﺴﺘﻁﻴﻊ‬
‫ﺍﻟﻘﻴﺎﻡ ﺒﺎﻟﻤﻬﺎﻡ ﺍﻟﺘﻲ ﻴﻘﻭﻡ ﺒﻬﺎ ﺍﻹﻨﺴﺎﻥ ﻋﻠﻰ ﻨﺤﻭ ﺃﺩﻕ ﻭﺃﺴﺭﻉ ﻭﺒﻜﻠﻔﺔ ﺍﻗل ﻤﻤﺎ ﻴﻘﻭﻡ ﺒﻪ ﺍﻟﺒﺸﺭ ﻤـﻥ‬
‫ﺍﺠل ﻤﻭﺍﻜﺒﺔ ﺍﻟﺘﻁﻭﺭ ﺍﻟﻬﺎﺌل ﻭﺍﻟﺴﺭﻴﻊ ﻓﻲ ﺃﻨﺸﻁﺔ ﺍﻟﺤﻴﺎﺓ ﺍﻟﻤﺨﺘﻠﻔﺔ ]‪.[6‬‬
‫ﻭﻤﻥ ﺍﻟﻤﻼﺤﻅ ﺃﻥ ﺍﻷﻁﻔﺎل ﻓﻲ ﺴﻥ ﺍﻟﺨﺎﻤﺴﺔ ﻤﻥ ﺍﻟﻌﻤـﺭ ﻴـﺴﺘﻁﻴﻌﻭﻥ ﺘﻤﻴﻴـﺯ ﺍﻷﺭﻗـﺎﻡ‬
‫ﻭﺍﻟﺤﺭﻭﻑ ﺍﻤﺎ ﺍﻟﺤﺭﻭﻑ ﺍﻟﻜﺒﻴﺭﺓ ﻭﺍﻟﺼﻐﻴﺭﺓ ﻭﺍﻟﻤﻜﺘﻭﺒﺔ ﺒﺨﻁ ﺍﻟﻴـﺩ ﺃﻭ ﺍﻟﻤﻁﺒﻭﻋـﺔ ﻓﻴـﺴﺘﻁﻴﻊ ﺃﻥ‬
‫ﻴﻤﻴﺯﻫﺎ ﺍﻟﻜﺒﺎﺭ ‪ ،‬ﻭﻟﻘﺩ ﺤﺎﻭل ﺍﻹﻨﺴﺎﻥ ﺘﺩﺭﻴﺠﻴﺎ ﺘﻌﻠﻴﻡ ﺍﻻﻟﺔ ﺘﻤﻴﻴﺯ ﺍﻷﺸﻴﺎﺀ ﺍﻟﻤﺨﺘﻠﻔﺔ ﻟﻴﻅﻬﺭ ﻓﻴﻤﺎ ﺒﻌﺩ‬
‫ﺍﻟﺘﻤﻴﻴﺯ ﺍﻻﻟﻲ ﺃﻭ ﻤﺎ ﻴﻌﺭﻑ ﺒﺘﻤﻴﻴﺯ ﺍﻷﻨﻤﺎﻁ )‪ ، (Patterns recognition‬ﻭﻫﻭ ﺍﻟﻤﺠـﺎل ﺍﻟـﺫﻱ‬
‫ﻴﻌﻨﻰ ﺒﺩﺭﺍﺴﺔ ﻜﻴﻔﻴﺔ ﻗﻴﺎﻡ ﺍﻻﻟﺔ ﺒﻤﺭﺍﻗﺒﺔ ﺍﻟﺒﻴﺌﺔ ﻭﺘﻌﻠﻴﻤﻬﺎ ﺘﻤﻴﻴﺯ ﺍﻷﻨﻤﺎﻁ ﺫﺍﺕ ﺍﻻﻫﺘﻤﺎﻡ ﻤﻤﺎ ﻴﺤﻴﻁ ﺒﻬﺎ‬
‫ﺜﻡ ﺍﺘﺨﺎﺫ ﺍﻟﻘﺭﺍﺭ ﻟﺘﺼﻨﻴﻑ ﻫﺫﻩ ﺍﻷﻨﻤﺎﻁ ﺇﻟﻰ ﺃﺼﻨﺎﻑ‪ ،‬ﻭﺍﻷﻨﻤﺎﻁ ﺠﻤﻊ ﻨﻤﻁ ﻭﻫﻭ ﻴﻤﺜل ﻜـل ﻜـﺎﺌﻥ‬
‫ﻴﺄﺨﺫ ﻗﻴﻤﺔ ﻭﺍﺴﻤﺎ ﻜﺄﻥ ﻴﻜﻭﻥ ﺼﻭﺭﺓ ﻟﺒﺼﻤﺔ ﺇﺼﺒﻊ ﺃﻭ ﻜﻠﻤﺔ ﻤﻜﺘﻭﺒﺔ ﺒﺨﻁ ﺍﻟﻴﺩ ﺃﻭ ﻭﺠﻪ ﺇﻨـﺴﺎﻥ ﺃﻭ‬
‫ﻏﻴﺭ ﺫﻟﻙ ﺍﺫ ﻻ ﻴﺯﺍل ﺍﻟﺒﺎﺤﺜﻭﻥ ﻴﻘﺩﻤﻭﻥ ﺍﻟﻌﺩﻴﺩ ﻤﻥ ﺍﻟﺩﺭﺍﺴﺎﺕ ﻭﺍﻷﻓﻜﺎﺭ ﺍﻟﺠﺩﻴﺩﺓ ﻹﻴﺠﺎﺩ ﺍﻟﺘﻁﺒﻴﻘـﺎﺕ‬
‫ﻓﻲ ﻤﺠﺎل ﺘﻤﺜﻴل ﺍﻟﻤﻌﺎﻟﻡ ﻭﺘﻤﻴﻴﺯﻫﺎ ﺁﻟﻴﺎ ﻭﻴﻌﺩ ﺍﻟﺤﺎﺴﻭﺏ ﺍﻷﺴﺭﻉ ﻭﺍﻷﻜﺜﺭ ﺍﺴﺘﺨﺩﺍﻤﺎ ﻓـﻲ ﻋﻤﻠﻴـﺎﺕ‬
‫ﺍﻟﺘﺼﻨﻴﻑ ﻭﻗﺩ ﺍﺼﺒﺢ ﺍﺴﺘﺨﺩﺍﻡ ﺍﻷﻨﻅﻤﺔ ﺍﻟﻤﺭﺌﻴﺔ ﺍﻷﺴﻠﻭﺏ ﺍﻷﻜﺜﺭ ﺍﻨﺘﺸﺎﺭﺍ ﺍﺫ ﺘﺴﺘﺨﺩﻡ ﺍﻟﺼﻭﺭ ﻓـﻲ‬
‫ﺘﻤﺜﻴل ﺍﻟﺒﻴﺎﻨﺎﺕ ﻭﺫﻟﻙ ﻻﻥ ﻓﻲ ﺍﻤﻜﺎﻥ ﺍﻹﻨﺴﺎﻥ ﺃﻥ ﻴﻤﻴﺯ ﺍﻟﺼﻭﺭﺓ ﻭﻴﺘﺫﻜﺭﻫﺎ ﺃﻜﺜﺭ ﻤﻥ ﺍﻟﻜﺘﺎﺒﺔ ﻓﻀﻼ‬
‫ﻋﻥ ﻤﺴﺎﻋﺩﺘﻪ ﻓﻲ ﺍﻟﻔﻬﻡ ﻭﺍﻹﺩﺭﺍﻙ ﻋﻠﻰ ﻨﺤﻭ ﺃﺴﺭﻉ‪ .‬ﻟﺫﺍ ﻅﻬﺭﺕ ﺍﻟﺤﺎﺠﺔ ﺍﻟـﻰ ﺘﺤﻠﻴـل ﺍﻟـﺼﻭﺭﺓ‬
‫ﻻﺴﺘﺨﻼﺹ ﻤﻌﺎﻟﻤﻬﺎ ﻭﺘﺤﺩﻴﺩ ﺍﻷﺸﻜﺎل ﻓﻴﻬﺎ ﻭﺘﺤﻭﻴﻠﻬﺎ ﺇﻟﻰ ﺒﻴﺎﻨﺎﺕ ﺜﻡ ﻤﻌﺎﻟﺠﺘﻬﺎ ]‪.[10‬‬

‫‪90‬‬
‫‪90‬‬
 ‫ﺍﺴﺘﺨﺩﺍﻡ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﺍﻟﺠﻴﻨﻴﺔ ﻓﻲ ﺘﻁﺎﺒﻕ ‪....‬‬

‫‪ .6‬ﺘﻁﺒﻴﻘﺎﺕ ﺘﻤﻴﻴﺯ ﺍﻷﻨﻤﺎﻁ‬

‫ﺍﺘﺴﻌﺕ ﻤﺴﺎﺤﺔ ﻤﺠﺎل ﺘﻁﺒﻴﻕ ﺍﻷﻨﻤﺎﻁ ﻟﺘﺸﻤل ﺘﻁﺒﻴﻘﺎﺕ ﻋﺩﻴﺩﺓ ﻓﻲ ﻤﺠﺎﻻﺕ ﺍﻟﺤﻴﺎﺓ ﺍﻟﻤﺨﺘﻠﻔﺔ‬
‫ﻤﺜل ﺍﻟﺘﻁﺒﻴﻘﺎﺕ ﺍﻟﻬﻨﺩﺴﻴﺔ ﺍﻟﻁﺒﻴﺔ ﻭﻤﺠﺎﻻﺕ ﺃﺨﺭﻯ ﻤﺜل ﺤﻘل ﺍﻟﺫﻜﺎﺀ ﺍﻻﺼـﻁﻨﺎﻋﻲ ) ‪Artificial‬‬
‫ـﺴﻭﻴﻕ‬
‫ـﻥ ﺍﻟﺘـ‬
‫ـﻀﻼ ﻋـ‬
‫ـﻥ ﺒﻌ ـﺩ )‪ (Remote sensing‬ﻓـ‬
‫ـﺸﻌﺎﺭ ﻋـ‬
‫‪ (intelligence‬ﻭ ﺍﻻﺴﺘـ‬
‫)‪ (Marketing‬ﻭﻨﻅﺭﺍ ﻟﻠﺘﻁﻭﺭ ﺍﻟﻬﺎﺌل ﻓﻲ ﻋﻠﻭﻡ ﺍﻟﺤﺎﺴﻭﺏ ﻓﻘﺩ ﺃﺼﺒﺤﺕ ﺍﻟﺘﺤﺩﻴﺎﺕ ﺍﻟﺘﻲ ﺘﻭﺍﺠـﻪ‬
‫ﻤﺼﻤﻤﻲ ﺃﻨﻅﻤﺔ ﺘﻤﻴﻴﺯ ﺍﻷﻨﻤﺎﻁ ﻜﺒﻴﺭﺓ ﺠﺩﺍ ﺍﺫ ﺃﺼﺒﺢ ﻓﻲ ﺍﻹﻤﻜﺎﻥ ﺍﻟﺘﻌﺎﻤل ﺒﺴﺭﻋﺔ ﻤﻊ ﻤﺠﻤﻭﻋﺎﺕ‬
‫ﻫﺎﺌﻠﺔ ﻤﻥ ﺍﻟﺒﻴﺎﻨﺎﺕ ﻤﻤﺎ ﺃﺩﻯ ﺇﻟﻰ ﺃﻥ ﺘﻜﻭﻥ ﺃﻨﻅﻤﺔ ﺘﻤﻴﻴﺯ ﺍﻷﻨﻤﺎﻁ ﺃﻜﺜﺭ ﺇﺤﻜﺎﻤﺎ ﻓﻲ ﺍﻟﺘﺼﻤﻴﻡ ﻟﺘﺤﻘﻕ‬
‫ﺃﻋﻠﻰ ﺍﻟﻤﺴﺘﻭﻴﺎﺕ ﻓﻲ ﺍﻷﺩﺍﺀ ﻤﻥ ﺤﻴﺙ ﺍﻟﺴﺭﻋﺔ ﻭﺍﻟﺩﻗﺔ ﻭﺍﻟﻜﻠﻔﺔ‪.‬‬
‫ﺇﻥ ﺘﺼﻤﻴﻡ ﺃﻱ ﻨﻅﺎﻡ ﻟﺘﻤﻴﻴﺯ ﺍﻷﻨﻤﺎﻁ ﻴﺘﻁﻠﺏ ﺍﻻﻨﺘﺒﺎﻩ ﺇﻟﻰ ﺍﻷﻤﻭﺭ ﺍﻻﺘﻴﺔ ]‪:[6‬‬
‫ﺘﻌﺭﻴﻑ ﺍﻷﺼﻨﺎﻑ ‪ ،‬ﺘﻤﺜﻴل ﺍﻟﻨﻤﻁ ‪ ،‬ﺍﺴﺘﺨﻼﺹ ﺍﻟﺨﺼﺎﺌﺹ‪ ،‬ﺘﺼﻤﻴﻡ ﺍﻟﻤﺼﻨﻑ ‪ ،‬ﺘﻘﻴﻴﻡ ﺍﻷﺩﺍﺀ‪.‬‬

‫‪ .7‬ﺍﻟﺠﺎﻨﺏ ﺍﻟﻌﻤﻠﻲ ﻟﻠﺒﺤﺙ‬

‫ﻴﻭﻀﺢ ﺍﻟﺒﺤﺙ ﺍﻟﺤﺎﻟﻲ ﻁﺭﻴﻘﺔ ﺘﻁﺎﺒﻕ ﻨﻤﻁ ﺍﻟﺤﺭﻑ ﺍﻹﻨﻜﻠﻴﺯﻱ ﺍﻷﺼﻴل ﺍﻟـﺫﻱ ﺍﺩﺨـل ﻤـﻊ‬
‫ﺃﺭﺒﻌﺔ ﺃﻨﻤﺎﻁ ﻤﺸﻭﻫﺔ )ﺒﻴﺎﻨﺎﺕ ﺍﻟﻨﻤﻁ ﻟﻠﺤﺭﻑ ﺍﻟﻤﺸﻭﻩ ﺍﻟﻤﺩﺨل ﻻﺘﺘﻁﺎﺒﻕ ﻤﻊ ﺒﻴﺎﻨﺎﺕ ﺍﻟﺤﺭﻑ ﺍﻻﺼﻴل‬
‫ﺍﻟﻤﺩﺨل( ﻤﻥ ﺤﻴﺙ ﺍﻟﺸﻜل ﻟﻠﺤﺭﻑ ﺍﻹﻨﻜﻠﻴﺯﻱ ﻨﻔﺴﻪ ﺍﻟﺫﻱ ﺍﺩﺨل ﺒﺎﻻﻋﺘﻤﺎﺩ ﻋﻠﻰ ﺍﺤﺩ ﺃﺴﺎﻟﻴﺏ ﺍﻟﺫﻜﺎﺀ‬
‫ﺍﻻﺼﻁﻨﺎﻋﻲ ﻭﻫﻲ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﺍﻟﺠﻴﻨﻴﺔ ﺫﻟﻙ ﺍﻥ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﺍﻟﺠﻴﻨﻴﺔ ﺘﻁﺒﻕ ﺒﻨﺠﺎﺡ ﻹﻴﺠﺎﺩ ﺍﻟﺤـل‬
‫ﺍﻟﻘﺭﻴﺏ ﻤﻥ ﺍﻟﻤﺜﺎﻟﻲ ﻓﻲ ﺍﻟﻤﺴﺎﺌل ﺍﻟﻤﺘﻌﻠﻘﺔ ﺒﺘﻤﻴﻴﺯ ﺍﻷﻨﻤﺎﻁ ﻜﻤﺎ ﻭﺍﻥ ﺍﻟﺒﺭﻨﺎﻤﺞ ﻴﻤﺘﻠﻙ ﻤﺭﻭﻨـﺔ ﻓـﻲ‬
‫ﺘﺤﺩﻴﺩ ﻋﺩﺩ ﺍﻻﻨﻤﺎﻁ ﺍﻟﻤﺸﻭﻫﺔ ﻭﺒﺎﻻﻤﻜﺎﻥ ﺍﺩﺨﺎل ﺍﻜﺜﺭ ﻤﻥ ﺍﺭﺒﻌﺔ ﺍﻨﻤﺎﻁ ﻤﺸﻭﻫﺔ ﻟﻠﺤﺭﻑ ﺍﻻﻨﻜﻠﻴﺯﻱ‬
‫ﻟﺫﻟﻙ ﻓﺎﻥ ﺴﺘﺭﺍﺘﻴﺠﻴﺔ ﺍﻟﺘﻨﻔﻴﺫ ﻟﻬﺫﻩ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﺘﻜﻭﻥ ﻋﻠﻰ ﻤﺭﺤﻠﺘﻴﻥ‪:‬‬
‫ﺍﻟﻤﺭﺤﻠﺔ ﺍﻷﻭﻟﻰ ‪ :‬ﺘﺘﻀﻤﻥ ﺍﻟﻤﺭﺤﻠﺔ ﺍﻷﻭﻟﻰ ﻁﻭﺭ ﺍﻟﺘﻌﻠﻴﻡ ﺍﺫ ﻴـﺩﺨل ﻨﻤـﻁ ﺍﻟﺤـﺭﻑ ﺍﻹﻨﻜﻠﻴـﺯﻱ‬
‫ﺍﻷﺼﻴل ﻏﻴﺭ ﺍﻟﻤﺸﻭﻩ ﻋﻠﻰ ﺸﻜل ﻤﺼﻔﻭﻓﺔ ﺫﺍﺕ ﺒﻌﺩﻴﻥ ‪ 8×8‬ﺍﺫ ﺘﺤﺘﻭﻱ ﻋﻠﻰ ‪ 64‬ﺨﻠﻴﺔ ﺫﺍﺕ ﻗـﻴﻡ‬
‫ﻤﻤﺜﻠﺔ ﺒﺎﻟﻭﺍﺤﺩ ﻭﺍﻟﺼﻔﺭ ‪،‬ﻭﻴﻤﺜل ﺍﻟـ ‪ bit‬ﺍﻟﺫﻱ ﻴﺤﻤل ﺍﻟﻘﻴﻤﺔ ﻭﺍﺤﺩ ﺒﺎﻟﻠﻭﻥ ﺍﻷﺴﻭﺩ ﻭﻴﻤﺜل ﺍﻟـ ‪bit‬‬
‫ﺍﻟﺫﻱ ﻴﺤﻤل ﺍﻟﻘﻴﻤﺔ ﺼﻔﺭ ﺒﺎﻟﻠﻭﻥ ﺍﻷﺒﻴﺽ ﻭﻫﻜﺫﺍ ﻟﺫﻟﻙ ﻓﺎﻥ ﺒﻴﺎﻨﺎﺕ ﺍﻟﻨﻤﻁ ﻟﻠﺤﺭﻑ ﺍﻹﻨﻜﻠﻴﺯﻱ ﺴﻭﻑ‬
‫ﺘﻤﺜل ﺒﻘﻴﻡ ﺜﻨﺎﺌﻴﺔ )‪ ، (Binary value‬ﺜﻡ ﺘﺤﻭل ﺍﻟﻘﻴﻡ ﺍﻟﺜﻨﺎﺌﻴﺔ ﺇﻟﻰ ﻗﻴﻡ ﻤﻤﺜﻠﺔ ﺒﻨﻅﺎﻡ ﺍﻟﺴﺎﺩﺱ ﻋﺸﺭ‬
‫)‪ (Hexadecimal‬ﻟﻐﺭﺽ ﺴﻬﻭﻟﺔ ﺍﻟﺒﺭﻤﺠﺔ ﻋﻨﺩ ﺇﺠﺭﺍﺀ ﻋﻤﻠﻴﺔ ﺍﻟﺘﻘﺎﻁﻊ ﻭﺍﻟﻁﻔﺭﺓ ﺍﻟﻭﺭﺍﺜﻴﺔ ﺜﻡ ﺘﺘﻡ‬
‫ﻋﻤﻠﻴﺔ ﺤﺴﺎﺏ ﻗﻴﻤﺔ ﺩﺍﻟﺔ ﺍﻟﻠﻴﺎﻗﺔ ﻟﻠﻨﻤﻁ ﺍﻷﺼﻴل ﻋﻥ ﻁﺭﻴﻕ ﺍﻟﻤﻌﺎﺩﻟﺔ ﺍﻻﺘﻴﺔ ‪:‬‬
‫) ‪fit ( x) = ∑ ( I x ∗ K y‬‬

‫‪ : I‬ﻴﻤﺜل ﺭﻗﻡ ﺃﺴﻁﺭ ﺍﻟﻤﺼﻔﻭﻓﺔ ﻭﺘﺘﻜﻭﻥ ﻤﻥ ﺜﻤﺎﻨﻴﺔ ﺍﺴﻁﺭ ‪.1<=x<=8‬‬ ‫‪x‬‬

‫‪ : K Y‬ﺘﻤﺜل ﻤﺤﺘﻭﻴﺎﺕ ﻜل ﺴﻁﺭ ﻤﻥ ﺍﻟﻤﺼﻔﻭﻓﺔ ﻭﺘﺘﻀﻤﻥ ﺍﻟﻘﻴﻡ ﺍﻟﺜﻨﺎﺌﻴﺔ ﺍﻟﺘﻲ ﺤﻭﻟﺕ ﺇﻟﻰ ﻨﻅﺎﻡ‬
‫ﺍﻟﺴﺎﺩﺱ ﻋﺸﺭ ‪.1<=y<=8‬‬

‫‪91‬‬
‫‪91‬‬
 ‫ﻤﻬﺎ ﻋﺒﺩ ﺍﻹﻟﻪ ﻤﺤﻤﺩ ﺍﻟﺒﺩﺭﺍﻨﻲ‬

‫ﻓﻔﻲ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﺍﻟﺠﻴﻨﻴﺔ ﻴﺘﻡ ﺘﺨﺼﻴﺹ ﻗﻴﻤﺔ ﺼﻼﺤﻴﺔ ) ‪ (fitness value‬ﻟﻜل ﻓـﺭﺩ‬
‫ﺴﻭﺍﺀ ﻜﺎﻥ ﻤﻭﻟﺩﺍ ﻓﻲ ﺍﻟﻤﺠﺘﻤﻊ ﺍﻻﺒﺘﺩﺍﺌﻲ ﺍﻭ ﻤﻥ ﺨﻼل ﺍﻻﺠﻴﺎل ﺍﻟﻼﺤﻘﺔ ﻭﺘﺭﺘﺒﻁ ﻫﺫﻩ ﺍﻟﻘﻴﻤﺔ ﺒﻘﻴﻤـﺔ‬
‫ﺩﺍﻟﺔ ﺍﻟﻬﺩﻑ )‪ (objective function‬ﻓﻤﺜ ﹰ‬
‫ﻼ ﺘﻜﻭﻥ ﻗﻴﻤﺔ ﺍﻟﺼﻼﺤﻴﺔ ﻟﻤﺴﺎﺌل ﺍﻟﺘﻌﻅﻴﻡ ﻤـﺴﺎﻭﻴﺔ‬
‫ﻟﻘﻴﻤﺔ ﺩﺍﻟﺔ ﺍﻟﻬﺩﻑ‪ ،‬ﺍﻤﺎ ﻤﺴﺎﺌل ﺍﻟﺘﺼﻐﻴﺭ ﻓﺄﻥ ﺍﻟﻬﺩﻑ ﻫﻭ ﺍﻴﺠﺎﺩ ﺍﻟﺤل ﺫﻱ ﺍﻟﻘﻴﻤﺔ ﺍﻟﺩﻨﻴﺎ ﻟﺩﺍﻟﺔ ﺍﻟﻬﺩﻑ‬
‫ﻭﺒﺫﻟﻙ ﺘﺤﺼل ﺍﻟﺤﻠﻭل ﺫﻭﺍﺕ ﺍﻟﻘﻴﻡ ﺍﻟﺼﻐﺭﻯ ﻟﺩﻭﺍل ﺍﻟﻬﺩﻑ ﻋﻠﻰ ﻗﻴﻤﺔ ﺍﻜﺒﺭ ﻟﻠﺼﻼﺤﻴﺔ‪.‬‬
‫ﺜﻡ ﺘﺨﺯﻥ ﻗﻴﻤﺔ ﺩﺍﻟﺔ ﺍﻟﻠﻴﺎﻗﺔ ﻟﻜل ﺤﺭﻑ ﻓﻲ ﻤﻠﻑ ﻻﻥ ﻫﺫﻩ ﺍﻟﻘﻴﻤﺔ ﺘﻌﺩ ﻗﻴﻤﺔ ﺩﺍﻟـﺔ ﺍﻟﻠﻴﺎﻗـﺔ‬
‫ﺍﻟﻤﺜﺎﻟﻴﺔ ﻟﻜل ﺤﺭﻑ ﻤﻥ ﺍﻟﺤﺭﻭﻑ ﺍﻹﻨﻜﻠﻴﺯﻴﺔ ﺍﻷﺼﻴﻠﺔ‪ ،‬ﻭﺍﻟﺸﻜل )‪ (6‬ﻴﻤﺜل ﻭﺍﺠﻬﺔ ﻁﻭﺭ ﺍﻟﺘﻌﻠﻴﻡ ﻓﻲ‬
‫ﺍﻟﺒﺭﻨﺎﻤﺞ ﺍﺫ ﺍﺩﺨل ﺤﺭﻑ ‪ B‬ﻤﺜﺎ ﹰ‬
‫ﻻ ﻋﻠﻰ ﻤﺎ ﺫﻜﺭ ﻓﻲ ﺃﻋﻼﻩ ‪.‬‬

‫ﺍﻟﺸﻜل )‪ (6‬ﻴﻤﺜل ﻭﺍﺠﻬﺔ ﻁﻭﺭ ﺍﻟﺘﻌﻠﻴﻡ‬

‫ﺍﻟﻤﺭﺤﻠﺔ ﺍﻟﺜﺎﻨﻴﺔ ‪ :‬ﺘﺘﻀﻤﻥ ﻫﺫﻩ ﺍﻟﻤﺭﺤﻠﺔ ﺇﺩﺨﺎل ﺃﺭﺒﻌﺔ ﺃﻨﻤﺎﻁ )ﺃﺒﺎﺀ( ﻤـﺸﻭﻫﻴﻥ ﻴﻤﺜﻠـﻭﻥ‬
‫ﺍﻟﺠﻴل ﺍﻻﺒﺘﺩﺍﺌﻲ ﺒﻁﺭﻴﻘﺔ ﺇﺩﺨﺎل ﻨﻤﻁ ﺍﻟﺤﺭﻑ ﺍﻹﻨﻜﻠﻴﺯﻱ ﺍﻷﺼﻴل ﻨﻔﺴﻬﺎ ﺜﻡ ﺘﺠﺭﻯ ﻋﻤﻠﻴﺔ ﺤـﺴﺎﺏ‬
‫ﻗﻴﻤﺔ ﺩﺍﻟﺔ ﺍﻟﻠﻴﺎﻗﺔ ﻟﻜل ﻨﻤﻁ ﺒﻁﺭﻴﻘﺔ ﺤﺴﺎﺏ ﻗﻴﻤﺔ ﺩﺍﻟﺔ ﺍﻟﻠﻴﺎﻗﺔ ﻟﻠﻨﻤﻁ ﺍﻷﺼﻴل ﻨﻔﺴﻬﺎ ﻭﺜـﻡ ﺘﺤـﺴﺏ‬
‫ﺍﺤﺘﻤﺎﻟﻴﺔ ﺍﻟﻤﺴﺎﻫﻤﺔ ﻟﻜل ﻨﻤﻁ ‪.‬‬
‫ﻭﺍﻟﺸﻜل)‪ (7‬ﻴﻭﻀﺢ ﺃﺭﺒﻌﺔ ﻨﻤﺎﺫﺝ ﻤﻥ ﺍﻷﻨﻤﺎﻁ ﺍﻟﻤﺸﻭﻫﺔ ﻟﻠﺤﺭﻑ ‪ B‬ﺍﺩﺨﻠﺕ ﺇﻟﻰ ﺍﻟﺒﺭﻨﺎﻤﺞ‬
‫ﻭﺘﻤﺜل ﺍﻟﺠﻴل ﺍﻻﺒﺘﺩﺍﺌﻲ ‪.‬‬

‫‪92‬‬
‫‪92‬‬
 ‫ﺍﺴﺘﺨﺩﺍﻡ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﺍﻟﺠﻴﻨﻴﺔ ﻓﻲ ﺘﻁﺎﺒﻕ ‪....‬‬

‫ﺍﻟﺸﻜل )‪ (7‬ﻴﻤﺜل ﻨﻤﺎﺫﺝ ﻤﻥ ﺍﻷﻨﻤﺎﻁ ﺍﻟﻤﺸﻭﻫﺔ ﻟﻠﺤﺭﻑ‪B‬‬

‫ﺒﻌﺩ ﺫﻟﻙ ﺘﺤﺩﺙ ﻋﻤﻠﻴﺔ ﺍﺨﺘﻴﺎﺭ ﺍﻷﺒﺎﺀ )‪ (parents selection function‬ﻤـﻥ ﺍﺠـل‬
‫ﺍﻨﺘﻘﺎﺀ ﻜﺭﻭﻤﻭﺴﻭﻤﺎﺕ ﻤﻥ ﺍﻟﻤﺠﺘﻤﻊ ﻟﺘﻜﻭﻴﻥ ﺃﺒﺎﺀ ﻟﻌﻤﻠﻴﺔ ﺍﻟﺘﻘﺎﻁﻊ ﻭﻹﻨﺘﺎﺝ ﺠﻴل ﺠﺩﻴﺩ ﻭﻗﺩ ﺍﺴﺘﺨﺩﺍﻤﺕ‬
‫ﻁﺭﻴﻘﺔ ﺍﻻﻨﺘﻘﺎﺀ ﺍﻟﻨﺴﺒﻲ )‪ (Proportional selection‬ﻻﺨﺘﻴﺎﺭ ﺍﻷﺒﺎﺀ ﻭﺒﺤﺴﺏ ﺍﻟﻤﻌﺎﺩﻟﺔ ﺍﻻﺘﻴﺔ ‪:‬‬
‫)) ‪f ( xi(t‬‬
‫= )‪prop( x‬‬
‫)) ‪∑ f ( xi(t‬‬

‫ﻭﺍﻟﺸﻜل)‪ (8‬ﻴﻭﻀﺢ ﺸﻜل ﺍﻷﻨﻤﺎﻁ ﺍﻟﻤﺸﻭﻫﺔ ﺍﻟﻤﺩﺨﻠﺔ ﻟﻠﺤـﺭﻑ ‪ B‬ﺍﻟﺘـﻲ ﺍﺩﺨﻠـﺕ ﺇﻟـﻰ‬
‫ﺍﻟﺒﺭﻨﺎﻤﺞ ﻭﺍﻟﺘﻲ ﺘﻤﺜل ﺍﻟﺠﻴل ﺍﻻﺒﺘﺩﺍﺌﻲ ‪.‬‬

‫‪93‬‬
‫‪93‬‬
 ‫ﻤﻬﺎ ﻋﺒﺩ ﺍﻹﻟﻪ ﻤﺤﻤﺩ ﺍﻟﺒﺩﺭﺍﻨﻲ‬

‫ﺍﻟﺸﻜل)‪ (8‬ﻴﻭﻀﺢ ﺸﻜل ﺍﻷﻨﻤﺎﻁ ﺍﻟﻤﺸﻭﻫﺔ ﺍﻟﻤﺩﺨﻠﺔ ﻟﻠﺤﺭﻑ ‪B‬‬

‫ﺒﻌﺩ ﺫﻟﻙ ﺘﺤﺩﺙ ﻋﻤﻠﻴﺔ ﺍﻟﺘﻘـﺎﻁﻊ )‪ (Crossover‬ﺍﺫ ﻨـﺩﻤﺞ ﻜﺭﻭﻤﻭﺴـﻭﻤﻴﻥ ﻟﺘﻜـﻭﻴﻥ‬

‫ﻜﺭﻭﻤﻭﺴﻭﻡ ﺠﺩﻴﺩ ﺃﻱ ﺍﺨﺘﻴﺎﺭ ﻓﺭﺩﻴﻥ )ﻨﻤﻁﻴﻥ( ﻤﻥ ﺍﻟﻤﺠﺘﻤﻊ ﻟﻠﺘﻘﺎﻁﻊ ﻭﻗـﺩ ﺍﺴـﺘﺨﺩﺍﻤﺕ ﻁﺭﻴﻘـﺔ‬
‫ﺍﻟﺘﻘﺎﻁﻊ ﺫﻱ ﻨﻘﻁﺘﻲ ﺍﻟﻘﻁﻊ )‪ (Two – point crossover‬ﻭﺍﻟﻔﻜﺭﺓ ﻤﻥ ﻫﺫﻩ ﺍﻟﻌﻤﻠﻴـﺔ ﻫـﻲ ﺍﻥ‬
‫ﺍﻟﻜﺭﻭﻤﻭﺴﻭﻡ ﺍﻟﻨﺎﺘﺞ ﻴﻜﻭﻥ ﺍﻓﻀل ﻤﻥ ﺍﻷﺒﻭﻴﻥ ﺍﺫﺍ ﺍﺨﺫ ﺍﻟﺼﻔﺎﺕ ﺍﻟﺠﻴﺩﺓ ﻤﻥ ﻜﻠﻴﻬﻤﺎ‪ ،‬ﺜﻡ ﺘﻅﻬﺭ ﻗـﻴﻡ‬
‫ﺍﻟﺠﻴﻥ ﻟﻠﻜﺭﻭﻤﻭﺴﻭﻡ ﺍﻟﻨﺎﺘﺞ ﻤﻥ ﺨﻼل ﺃﻴﺠﺎﺩ ﺩﺍﻟﺔ ﺍﻟﻁﻔﺭﺓ ﺍﻟﻭﺭﺍﺜﻴﺔ ﻀﻤﻥ ﺍﻟﻤﻘﻁـﻊ ‪(Mutation‬‬
‫)‪ function‬ﻭﻗﺩ ﺍﺴﺘﺨﺩﺍﻤﺕ ﺼﻴﻐﺔ ﺍﻟﻁﻔﺭﺓ ﺍﻟﻭﺭﺍﺜﻴﺔ ﺒﺕ )‪ (Bit‬ﻷﻨﻬﺎ ﺘﺴﺘﻌﻤل ﻤـﻊ ﺍﻟﺘـﺸﻔﻴﺭ‬
‫ﺍﻟﺜﻨﺎﺌﻲ ‪ ،‬ﻭﺒﻌﺩ ﺫﻟﻙ ﻨﻭﺠﺩ ﺩﺍﻟﺔ ﺍﻟﻬﺩﻑ ﻟﻤﻘﺎﻁﻊ ﺍﻟﺠﻴل ﺍﻟﺠﺩﻴﺩ ‪ ،‬ﺜﻡ ﻨﻭﺠﺩ ﺩﺍﻟﺔ ﺍﻟﻠﻴﺎﻗـﺔ ﻭﺍﺤﺘﻤﺎﻟﻴـﺔ‬
‫ﺍﻟﻤﺴﺎﻫﻤﺔ ﻟﻤﻘﺎﻁﻊ ﺍﻟﺠﻴل ﺍﻟﺠﺩﻴﺩ ‪.‬‬
‫ﻭﺍﻟﺸﻜل)‪ (9‬ﻴﻭﻀﺢ ﻗﻴﻡ ﺍﻷﻨﻤﺎﻁ ﺍﻟﻤﺸﻭﻫﺔ ﺍﻟﻨﺎﺘﺠﺔ ﻟﻠﺤﺭﻑ ‪ B‬ﺒﻌﺩ ﺇﺠﺭﺍﺀ ﻋﻤﻠﻴﺘﻲ ﺍﻟﺘﻘﺎﻁﻊ‬
‫ﻭﺍﻟﻁﻔﺭﺓ ﺍﻟﻭﺭﺍﺜﻴﺔ ﻭﻫﺫﻩ ﺍﻷﻨﻤﺎﻁ ﺘﻤﺜل ﺍﻟﺠﻴل ﺍﻷﻭل ﺍﻟﺠﺩﻴﺩ ‪.‬‬

‫‪94‬‬
‫‪94‬‬
 ‫ﺍﺴﺘﺨﺩﺍﻡ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﺍﻟﺠﻴﻨﻴﺔ ﻓﻲ ﺘﻁﺎﺒﻕ ‪....‬‬

‫ﺍﻟﺸﻜل)‪ (9‬ﻴﻭﻀﺢ ﻗﻴﻡ ﺍﻷﻨﻤﺎﻁ ﺍﻟﻤﺸﻭﻫﺔ ﺍﻟﻨﺎﺘﺠﺔ‬

‫ﻴﻭﻀﺢ ﺍﻟﺸﻜل)‪ (10‬ﺍﺸﻜﺎل ﺍﻷﻨﻤﺎﻁ ﺍﻟﻤﺸﻭﻫﺔ ﺍﻟﻨﺎﺘﺠﺔ ﻟﻠﺤﺭﻑ ‪ B‬ﺒﻌﺩ ﺇﺠـﺭﺍﺀ ﻋﻤﻠﻴﺘـﻲ‬
‫ﺍﻟﺘﻘﺎﻁﻊ ﻭﺍﻟﻁﻔﺭﺓ ﺍﻟﻭﺭﺍﺜﻴﺔ ‪.‬‬

‫ﺍﻟﺸﻜل)‪ (10‬ﻴﻭﻀﺢ ﺍﺸﻜﺎل ﺍﻷﻨﻤﺎﻁ ﺍﻟﻤﺸﻭﻫﺔ‬

‫‪95‬‬
‫‪95‬‬
 ‫ﻤﻬﺎ ﻋﺒﺩ ﺍﻹﻟﻪ ﻤﺤﻤﺩ ﺍﻟﺒﺩﺭﺍﻨﻲ‬

‫ﻨﻼﺤﻅ ﺍﻥ ﺍﻷﻨﻤﺎﻁ ﺍﻟﻨﺎﺘﺠﺔ ﺍﻟﺘﻲ ﺘﻤﺜل ﺍﻟﺠﻴل ﺍﻷﻭل ﺍﻟﺠﺩﻴـﺩ )‪ (Parent1‬ﻭ)‪(Parent2‬ﻭ‬

‫)‪ (Parent3‬ﻭ)‪ (Parent4‬ﺘﺤﻤل ﻗﻴﻤﺔ ﺩﺍﻟﺔ ﺍﻟﻠﻴﺎﻗﺔ )‪(2580‬ﻭ)‪ (4048‬ﻭ )‪(4028‬ﻭ)‪ (3748‬ﻋﻠﻰ‬
‫ﺍﻟﺘﻭﺍﻟﻲ ‪،‬ﻭﺒﻤﺎ ﺍﻥ ﻗﻴﻤﺔ ﺩﺍﻟﺔ ﺍﻟﻠﻴﺎﻗﺔ ﺍﻷﺼﻴﻠﺔ ﺍﻟﻤﺜﺎﻟﻴﺔ ﻟﻠﺤﺭﻑ ‪ B‬ﻫﻲ )‪ (4144‬ﻭﺍﻥ ﺩﺍﻟـﺔ ﺍﻟﻬـﺩﻑ‬
‫ﺍﻟﻤﺴﺘﺨﺩﻤﺔ ﺘﻤﺜل ﺩﺍﻟﺔ ﺘﻌﻅﻴﻡ )‪ (maximization function‬ﻓﺎﻥ )‪ (Parent2‬ﺴﻭﻑ ﻴﻜـﺭﺭ ﻭ‬
‫ﻻ ﺠﺩﻴﺩﺓ )ﻟﻡ ﻴﺴﺒﻕ ﺘﻜﻭﻴﻨﻬﺎ ﻓﻲ ﺍﻷﺠﻴﺎل ﺍﻟـﺴﺎﺒﻘﺔ( ﺴـﻭﻑ‬
‫)‪ (Parent1‬ﺴﻭﻑ ﻴﻬﻤل ﺫﻟﻙ ﺍﻥ ﺤﻠﻭ ﹰ‬
‫ﺘﻀﺎﻑ ﺍﻟﻰ ﺍﻟﻤﺠﺘﻤﻊ ﺍﻟﺠﺩﻴﺩ ﻓﻲ ﺤﺎﻟﺔ ﺤﺼﻭل ﺍﻻﻨﺘﻘﺎﺀ ﻭﺍﻟﺘﻘﺎﻁﻊ ﻭﺍﻟﻁﻔﺭﺓ ﺍﺫ ﺴﻭﻑ ﺘﻜـﻭﻥ ﻟﻬـﻡ‬
‫ﺍﺤﺘﻤﺎﻟﻴﺔ ﻤﺴﺎﻫﻤﺔ ﺃﻓﻀل ﻤﻥ ﺍﻷﻓﺭﺍﺩ ﺍﻟﺴﺎﺒﻘﻴﻥ ﻭﻫﺫﻩ ﺍﻟﻌﻤﻠﻴﺔ ﺘﻬﺩﻑ ﺍﻟﻰ ﺘﻭﺴـﻴﻊ ﻤﺠـﺎل ﺍﻟﺤﻠـﻭل‬
‫ﺍﻟﻤﻤﻜﻨﺔ ﻭﺍﻟﻤﺘﻤﺜل ﺒﺘﻜﻭﻴﻥ ﺍﻜﺒﺭ ﻋﺩﺩ ﻤﻥ ﺍﻷﻓﺭﺍﺩ ﺍﻟﺫﻴﻥ ﻴﻤﺘﻠﻜﻭﻥ ﺼﻔﺎﺕ ﻭﺭﺍﺜﻴﺔ ﻤﺭﻏـﻭﺏ ﻓﻴﻬـﺎ‬
‫ﻻ ﺇﻀﺎﻓﻴﺔ ﻤﻤﺎ ﻴﺅﺜﺭ ﺍﻴﺠﺎﺒﻴﹰﺎ ﻓﻲ ﺴﺭﻋﺔ ﺍﻗﺘﺭﺍﺏ ﺍﻟﺤل‪.‬‬
‫)ﺠﻴﺩﺓ( ﻀﻤﻥ ﺍﻟﻤﺠﺘﻤﻊ ﻤﻤﺎ ﻴﻭﻓﺭ ﺤﻠﻭ ﹰ‬

‫ﺃﻤﺎ ﻤﻌﻴﺎﺭ ﺍﻟﺘﻭﻗﻑ ﻓﻲ ﺍﻟﺒﺭﻨﺎﻤﺞ ﻓﻴﻌﺘﻤﺩ ﻋﻠﻰ ﺃﻥ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﺘﺘﻭﻗﻑ ﻋﻨﺩﻤﺎ ﺘﻜﻭﻥ ﺃﺤـﺴﻥ‬
‫ﻗﻴﻤﺔ ﻟﺩﺍﻟﺔ ﺍﻟﻠﻴﺎﻗﺔ ﺍﻜﺒﺭ ﻤﻥ ﻗﻴﻤﺔ ﺍﻟﻠﻴﺎﻗﺔ ﺍﻟﺘﻲ ﺤﺩﺩﺕ )ﻓﻲ ﺤﺎﻟﺔ ‪ (Maximize‬ﺜﻡ ﺘﺠـﺭﻱ ﻋﻤﻠﻴـﺔ‬
‫ﺭﺴﻡ ﺍﻷﻨﻤﺎﻁ ﺍﻟﺠﺩﻴﺩﺓ ﺍﻟﻨﺎﺘﺠﺔ ﻭﺍﻟﻘﺭﻴﺒﺔ ﻤﻥ ﺍﻟﻨﻤﻁ ﺍﻟﻤﺜﺎﻟﻲ ﺍﻟﻤﻁﻠﻭﺏ ﻟﻠﺤﺭﻑ ﺍﻹﻨﻜﻠﻴﺯﻱ ﺜﻡ ﺘﻭﻀﻊ‬
‫ﻤﺤل ﺍﻷﻓﺭﺍﺩ )ﺍﻷﻨﻤﺎﻁ ﺍﻟﺴﺎﺒﻘﺔ( ﻓﻲ ﺍﻟﻤﺠﺘﻤﻊ ‪.‬‬

‫ﻭﺍﻟﺸﻜل )‪ (11‬ﻴﻭﻀﺢ ﺍﻟﻤﺨﻁﻁ ﺍﻟﻌﺎﻡ ﻟﻠﺨﻭﺍﺭﺯﻤﻴﺔ ﺍﻟﺠﻴﻨﻴﺔ ﺍﻟﻤﺴﺘﺨﺩﻤﺔ ﻓﻲ ﻋﻤﻠﻴﺔ ﺘﻁﺎﺒﻕ‬

‫ﺃﻨﻤﺎﻁ ﺍﻟﺤﺭﻑ ﺍﻹﻨﻜﻠﻴﺯﻱ ‪.‬‬

‫‪96‬‬
‫‪96‬‬
 ‫ﺍﺴﺘﺨﺩﺍﻡ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﺍﻟﺠﻴﻨﻴﺔ ﻓﻲ ﺘﻁﺎﺒﻕ ‪....‬‬

‫ﺍﻟﺒﺩﺍﻴﺔ‬

‫ﺇﺩﺨﺎل ﺍﻟﻨﻤﻁ ﺍﻷﺼﻴل ﻟﻠﺤﺭﻑ ﺍﻹﻨﻜﻠﻴﺯﻱ ﻋﻠﻰ ﺸﻜل ﻗﻴﻡ ﺜﻨﺎﺌﻴﺔ‬

‫ﺍﺩﺨﺎل ﺍﺭﺒﻌﺔ ﺍﻨﻤﺎﻁ ﻤﺸﻭﻫﺔ ﻟﻠﺤﺭﻑ ﺍﻻﻨﻜﻠﻴﺯﻱ ﻨﻔﺴﻪ ﺍﻟﻤﺩﺨل ﻋﻠﻰ ﺸﻜل ﻗﻴﻡ ﺜﻨﺎﺌﻴﺔ‬

‫ﺍﻴﺠﺎﺩ ﺩﺍﻟﺔ ﺍﻟﻬﺩﻑ ﺍﻟﺨﺎﺼﺔ ﺒﺎﻟﻤﺴﺎﻟﺔ‬

‫ﺩﺍﻟﺔ ﺍﻨﺘﻘﺎﺀ )ﺍﺨﺘﻴﺎﺭ( ﺍﻻﺒﺎﺀ‬

‫ﺩﺍﻟﺔ ﺍﻟﺘﺩﺍﺨل ﺍﻻﺒﺩﺍﻟﻲ‬

‫ﺩﺍﻟﺔ ﺍﻟﺘﻐﻴﻴﺭ )ﺍﻟﻁﻔﺭﺓ ﺍﻟﻭﺭﺍﺜﻴﺔ(‬

‫ﺍﻟﻨﺘﺎﺌﺞ ﻏﻴﺭ ﻤﻨﺎﺴﺒﺔ‬

‫ﺍﻴﺠﺎﺩ ﺩﺍﻟﺔ ﺍﻟﻠﻴﺎﻗﺔ)‪ (fitness‬ﻟﻤﻘﺎﻁﻊ ﺍﻟﺠﻴل ﺍﻟﺠﺩﻴﺩ‬ ‫ﻟﻼﺴﺘﻤﺭﺍﺭ ﻓﻲ ﻋﻤﻠﻴﺔ‬
‫ﺍﻟﺘﺤﺴﻴﻥ‬
‫ﺍﻟﻨﺘﺎﺌﺞ ﻤﻨﺎﺴﺒﺔ‬

‫ﻤﻌﻴﺎﺭ ﺍﻟﺘﻭﻗﻑ ﻓﻲ‬

‫ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ‬

‫ﻁﺒﺎﻋﺔ ﺍﻟﺠﻴل ﺍﻟﻘﺭﻴﺏ ﻤﻥ ﺍﻟﻤﺜﺎﻟﻲ‬

‫ﺍﻟﻨﻬﺎﻴﺔ‬

‫ﺍﻟﺸﻜل)‪ (11‬ﺍﻟﻤﺨﻁﻁ ﺍﻟﻌﺎﻡ ﻟﻠﺨﻭﺍﺭﺯﻤﻴﺔ ﺍﻟﺠﻴﻨﻴﺔ ﺍﻟﻤﺴﺘﺨﺩﻤﺔ ﻓﻲ ﺘﻁﺎﺒﻕ ﺃﻨﻤﺎﻁ ﺍﻟﺤﺭﻑ‬

‫‪97‬‬
‫‪97‬‬
 ‫ﻤﻬﺎ ﻋﺒﺩ ﺍﻹﻟﻪ ﻤﺤﻤﺩ ﺍﻟﺒﺩﺭﺍﻨﻲ‬

‫‪ .8‬ﺍﻻﺴﺘﻨﺘﺎﺠﺎﺕ‬
‫ﺇﻥ ﻫﻨﺎﻟﻙ ﻋﺩﺓ ﺍﺴﺘﻨﺘﺎﺠﺎﺕ ﺘﻡ ﺍﻟﺘﻭﺼل ﺇﻟﻴﻬﺎ ﻤﻥ ﺃﻫﻤﻬﺎ ﺃﻥ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﺍﻟﺠﻴﻨﻴﺔ ﺘﻤﻜﻨﺕ ﻤﻥ‬
‫‪) Generation‬ﺘﻭﻟﻴﺩ ﻋﺩﺓ ﺃﺠﻴﺎل( ﻤﻥ ﺍﻷﻨﻤﺎﻁ ﺍﻟﻤﺴﺘﺨﺩﻤﺔ ﻭﺒﻁﺭﻴﻘﺔ ﻜﻔـﻭﺀﺓ‬ ‫ﺇﺠﺭﺍﺀ ﻋﻤﻠﻴﺔ ﺍﻟـ‬
‫ﻤﻥ ﺨﻼل ﺍﺴﺘﺨﺩﺍﻡ ﺃﺴﻠﻭﺏ ﻤﻁﺎﺒﻘﺔ ﺍﻟﻘﺎﻟﺏ ﻋﻨﺩ ﺇﺠﺭﺍﺀ ﻋﻤﻠﻴﺔ ﺘﻁﺎﺒﻕ ﻨﻤـﻁ ﺍﻟﺤـﺭﻑ ﺍﻹﻨﻜﻠﻴـﺯﻱ‬
‫ﺍﻷﺼﻴل ﻤﻊ ﺍﻷﻨﻤﺎﻁ ﺍﻟﻤﺸﻭﻫﺔ ﺍﻟﻤﺩﺨﻠﺔ ﺒﺎﻹﻀـﺎﻓﺔ ﺇﻟـﻰ ﺍﺴـﺘﺨﺩﺍﻡ ﻁﺭﻴﻘـﺔ ﺍﻻﻨﺘﻘـﺎﺀ ﺍﻟﻨـﺴﺒﻲ‬
‫)‪ (Proportional selection‬ﻭﺍﻟﺘﻘـﺎﻁﻊ ﺫﻭ ﻨﻘﻁﺘـﻲ ﺍﻟﻘﻁـﻊ )‪(Two – point crossover‬‬
‫ﻓﻀﻼ ﻋﻥ ﺃﻥ ﺍﻟﺠﻴـل ﺍﻷﻭل‬ ‫ﻭﺼﻴﻐﺔ ﺍﻟﻁﻔﺭﺓ ﺍﻟﻭﺭﺍﺜﻴﺔ ﺒﺕ )‪ (Bit‬ﻋﻨﺩ ﺘﻭﻟﻴﺩ ﺍﻷﺠﻴﺎل‪,‬‬
‫ﻼ ﺇﻟﻰ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﺍﻟﺠﻴﻨﻴﺔ ﺒﻌﺩ ﺘﻜﺭﺍﺭ ﺍﻟـﻨﻤﻁ ﺍﻟﻤﻘـﺎﺭﺏ ﻟـﻨﻤﻁ‬
‫ﻼ ﻤﺩﺨ ﹰ‬
‫ﺍﻟﺠﺩﻴﺩ ﺍﻟﻨﺎﺘﺞ ﻴﺼﺒﺢ ﺠﻴ ﹰ‬
‫ﺍﻟﺤﺭﻑ ﺍﻷﺼﻴل ﻭﺤﺫﻑ ﺍﻟﻨﻤﻁ ﻏﻴﺭ ﺍﻟﻤﻘﺎﺭﺏ ﻟﻨﻤﻁ ﺍﻟﺤﺭﻑ ﺍﻷﺼﻴل ﺒﺎﻻﻋﺘﻤﺎﺩ ﻋﻠﻰ ﻗﻴﻤـﺔ ﺩﺍﻟـﺔ‬
‫ﺍﻟﻠﻴﺎﻗﺔ ﻟﻠﺤﺭﻑ ﺍﻷﺼﻴل ﺍﻟﻤﺜﺎﻟﻲ ﺍﻟﺘﻲ ﺘﺨﺯﻥ ﻓﻲ ﻤﻠﻑ ﺨﺎﺹ ﻴﺘـﻀﻤﻥ ﻗـﻴﻡ ﺩﻭﺍل ﺍﻟﻠﻴﺎﻗـﺔ ﻟﻜـل‬
‫ﺍﻟﺤﺭﻭﻑ ﺍﻹﻨﻜﻠﻴﺯﻴﺔ ﺍﻷﺼﻴﻠﺔ ﻏﻴﺭ ﺍﻟﻤﺸﻭﻫﺔ ‪ ،‬ﻭﻗﺩ ﻻﺤﻅﻨﺎ ﺃﻥ ﻗﻴﻡ ﺩﺍﻟﺔ ﺍﻟﻠﻴﺎﻗﺔ ﻟﻠﺠﻴل ﺍﻟﺠﺩﻴﺩ ﺍﻟﻨﺎﺘﺞ‬
‫ﻼ ﻋﻥ ﺃﻥ ﺘﻤﺜﻴـل‬
‫ﺍﻓﻀل ﻤﻥ ﻗﻴﻡ ﺩﺍﻟﺔ ﺍﻟﻠﻴﺎﻗﺔ ﻟﻠﺠﻴل ﺍﻻﺒﺘﺩﺍﺌﻲ ﻓﻲ ﺤﺎﻟﺔ ﺘﻜﺭﺍﺭ ﺘﻭﻟﻴﺩ ﻷﺠﻴﺎل‪ ،‬ﻓﻀ ﹰ‬
‫ﺒﻴﺎﻨﺎﺕ ﺍﻟﻨﻤﻁ ﺒﻘﻴﻡ ﺜﻨﺎﺌﻴﺔ )‪ (binary value‬ﻭﻫﻲ ﺍﻟﺼﻔﺭ ﻭﺍﻟﻭﺍﺤﺩ ﻭﺘﺤﻭﻴﻠﻬﺎ ﺇﻟﻰ ﻗﻴﻡ ﻤﻤﺜﻠﺔ ﺒﻨﻅﺎﻡ‬
‫ﺍﻟﺴﺎﺩﺱ ﻋﺸﺭ ﺴﻬل ﻤﻥ ﺍﻟﺘﻤﺜﻴل ﻟﻠﻨﻤﻁ ﻭﺍﻟﺴﺭﻋﺔ ﻋﻨﺩ ﺘﻨﻔﻴﺫ ﻋﻤﻠﻴﺎﺕ ﺍﻻﻨﺘﻘﺎﺀ ﻭﺍﻟﺘﻘﺎﻁﻊ ﻭﺍﻟﻁﻔـﺭﺓ‬
‫ﺍﻟﻭﺭﺍﺜﻴﺔ ﻓﻲ ﺍﻟﺒﺭﻨﺎﻤﺞ‪.‬‬
‫ﻭﻴﻤﻜﻥ ﺍﻻﺴﺘﻤﺭﺍﺭ ﻓﻲ ﻋﻤﻠﻴﺔ ﺘﻭﻟﻴﺩ ﺠﻴل ﺠﺩﻴﺩ ﻤﻥ ﺃﻨﻤﺎﻁ ﺍﻟﺤﺭﻭﻑ ﺇﻟﻰ ﺃﻥ ﻴﺘﻡ ﺍﻟﺘﻭﻗـﻑ‬
‫ﻋﻥ ﻁﺭﻴﻕ ﻤﻌﻴﺎﺭ ﺍﻟﺘﻭﻗﻑ ﻓﻲ ﺍﻟﺒﺭﻨﺎﻤﺞ ﻭﺫﻟﻙ ﺒﺎﻻﻋﺘﻤﺎﺩ ﻋﻠﻰ ﻗﻴﻤﺔ ﺩﺍﻟﺔ ﺍﻟﻠﻴﺎﻗـﺔ ﻓـﻲ ﺍﻟﻤﺠﺘﻤـﻊ‬
‫ﺍﻟﺤﺎﻟﻲ ‪ ،‬ﻭﺘﺭﺒﻁ ﻫﺫﻩ ﺍﻟﻘﻴﻤﺔ ﺒﻘﻴﻤﺔ ﺩﺍﻟﺔ ﺍﻟﻬﺩﻑ ﻟﺫﻟﻙ ﻓﺎﻥ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﺍﻟﺠﻴﻨﻴﺔ ﺘﺴﺘﻌﻤل ﺩﺍﻟﺔ ﺍﻟﻬﺩﻑ‬
‫ﻤﺒﺎﺸﺭﺓ ﻭﻻ ﺘﺘﻭﺴﻊ ﻓﻲ ﻤﻌﻠﻭﻤﺎﺕ ﺇﻀﺎﻓﻴﺔ‪.‬‬
‫ﻭﺒﺎﻻﻤﻜﺎﻥ ﺘﻨﻔﻴﺫ ﺍﻟﺒﺤﺙ ﺍﻟﺤﺎﻟﻲ ﺒﺄﺴﺘﺨﺩﺍﻡ ﺃﻜﺜﺭ ﻤﻥ ﺃﺭﺒﻌﺔ ﻨﻤﺎﺫﺝ ﻜﻘﺎﻋﺩﺓ ﺒﻴﺎﻨﺎﺕ ﺒﺎﻟﻨـﺴﺒﺔ‬
‫ﻟﻠﺤﺭﻑ ﺍﻹﻨﻜﻠﻴﺯﻱ ﺍﻟﻤﺸﻭﻩ ﺍﻟﻤﺩﺨل ﺍﺫ ﺍﻥ ﺍﻟﺒﺭﻨﺎﻤﺞ ﻴﻤﺘﻠﻙ ﺍﻤﻜﺎﻨﻴﺔ ﺘﻁﺒﻴﻕ ﻋﻤﻠﻴﺔ ﺘﻤﻴﻴـﺯ ﺍﻻﻨﻤـﺎﻁ‬
‫ﻟﻠﺤﺭﻑ ﺍﻻﻨﻜﻠﻴﺯﻱ ﺍﻟﻤﺩﺨل ﻷﻜﺜﺭ ﻤﻥ ﺍﺭﺒﻌﺔ ﻨﻤﺎﺫﺝ ﻤﺸﻭﻫﺔ ﻭﺒﻁﺭﻴﻘﺔ ﻜﻔﻭﺀﺓ ﻻﻨﻪ ﻗﺩ ﺨﺯﻥ ﻗﻴﻤـﺔ‬
‫ﺩﺍﻟﺔ ﺍﻟﻠﻴﺎﻗﺔ ﻟﻜل ﺤﺭﻑ ﺍﺼﻴل ﻤﺩﺨل ﻓﻲ ﻤﻠﻑ ﺨﺎﺹ ﺒﺎﻻﻀﺎﻓﺔ ﺍﻟـﻰ ﺍﻤﻜﺎﻨﻴـﺔ ﺍﺴـﺘﺨﺩﺍﻡ ﻟﻐـﺔ‬
‫‪ Visual Basic‬ﺍﻟﺘﻲ ﺘﻌﺩ ﻤﻥ ﻟﻐﺎﺕ ﺍﻟﺒﺭﻤﺠﺔ ﺍﻟﺴﻬﻠﺔ ﺍﻻﺴﺘﺨﺩﺍﻡ ‪ ،‬ﻤﻤﺎ ﺴﻬل ﻟﻨﺎ ﺍﻟﺘﻌﺎﻤـل ﻤـﻊ‬
‫ﺍﻷﻨﻤﺎﻁ ﺍﻟﺘﻲ ﺘﻡ ﺘﻤﻴﻴﺯﻫﺎ ﻋﻥ ﻁﺭﻴﻕ ﺍﻟﺘﻌﺎﻤل ﻤﻊ ﺒﻴﺎﻨﺎﺕ ﻫﺫﻩ ﺍﻷﻨﻤﺎﻁ ‪.‬‬

‫‪98‬‬
‫‪98‬‬
 .... ‫ﺍﺴﺘﺨﺩﺍﻡ ﺍﻟﺨﻭﺍﺭﺯﻤﻴﺔ ﺍﻟﺠﻴﻨﻴﺔ ﻓﻲ ﺘﻁﺎﺒﻕ‬

‫ﺍﻟﻤﺼﺎﺩﺭ‬
Aaron Voegele, "Artificial Neural Network Feature Selection
Optimization with Genetic Algorithms", Department of General
Engineering , University of Illinois at Urbana-Champaign.
Andrew Deason, (2004) , Webmaster , “Genetic Algorithms” , Last
Modified
http://www.tjhsst.edu/~ai/AI2001/GA.HTM
Antionia J.J, (1993) : “Genetic Algorithms” and their application to the
design of neural Network , neural computing and Application , Springer-
Verlag , London Limited .
David E. Goldberg , (1993) , “Genetic Algorithm in search optimization
& machine learning “, the university of Alabama .
Hantao Zhang, (2003) , “Genetic Algorithms” , Linux Resources.
http://www.cs.toronto.edu/pub/xzhu/book/ai/geneticalg.pdf
J. T. Tou and R. C. Gonzalez , (1979) “Pattern Recognition Principles” ,
Addison – Wesley ,
http://www.csse.morash.edu.au/~sid/teach/cse3314/notesll.pdf
Kim-Fung Man. Kit-Sang Tang and Sam Kwong, (1998) : “ Genetic
Algorithms: Concepts and Designs”, city University of Hong Kong, IEEE
, Industrial.
Kluwer Ed. , (1997) : “Intelligent hybrid systems : Fuzzy Logic, Neural
networks and Genetic Algorithms , Kluwer Academic publishers ”.
Mitchell M., (1996) : “ An Introduction to Genetic Algorithms” ,
Abradford Book .
0. Pao. Yoh-Han , (1989) : “Adaptive Pattern Recognition and Neural
networks”, Publishing Company, Inc. , New york.

99
99