‫ﺍﻟﻤﻮﺯﻭﻧﺔ ﻟﺘﻘﺪﻳﺮ‬

‫ﺔ‬ ‫ﻁﺮﻳﻘﺔ ﺍﻟﻤﻤﺮﺑﻌﺎﺕ ﺍﻟﻟﺼﻐﺮﻯ ﺍ‬

‫ﺳﺘﺨﺪﺍﻡ ﻁ‬
‫ﺍﺳ‬

‫ﺳﻲ –‬
‫ﺨﺘﻠﻂ )ﺍﻻﺳ‬
‫ﻘﺎء ﻟﻠﻨﻤﻮﺫﺫﺝ ﺍﻻﺣﺘﻤﻤﺎﻟﻲ ﺍﻟﻤﺨ‬
‫ﺩﺍﻟﺔ ﺍﻟﺒﻘﺎ‬

‫ﻋﻠﻰ ﻋﻴﻨﺔ‬
‫ﻛﻛﺎﻣﺎ ﻣﻦ ﺍﻟﺮﺗﺒﺔ ﺍﻟﺜﺜﺎﻧﻴﺔ ( ﺩﺭﺭﺍﺳﺔ ﺗﻄﺒﺒﻴﻘﻴﺔ ﻰ‬

‫ﻣﻦ ﻣﻣﺮﺿﻰ ﺍﻟﻔﻔﺸﻞ ﺍﻟﻜﻠﻮﻮﻱ‬

‫ﻋﺑﺩ ﺍﻟﺣﺎﻓﻅ‬
‫ﻡ‪.‬ﺩ ﺍﺍﻳﻧﺎﺱ ﺩ‬
‫ﺏ‬
‫ﺻﺎﺣﺏ‬ ‫ﺣﺙ ﺍﺩﻫﻡ ﷴ‬
‫ﺍﻟﺑﺎﺣ‬
‫ﺣﺻﺎء‬
‫‪/‬ﺟﺎﻣﻌﻪ ﻛﺭﺑﻼء‪ -‬ﻗﺳﻡ ﺍﻻﺣ‬
‫ﻪ‬ ‫ﻛﻛﻠﻳﺔ ﺍﻹﺩﺍﺭﺭﺓ ﻭﺍﻻﻗﺗﺻ‬
‫ﺻﺎﺩ‬
 ‫ﺍﺳﺘﺨﺪﺍﻡ ﻁﺮﻳﻘﺔ ﺍﻟﻠﻤﺮﺑﻌﺎﺕ ﺍﻟﺼﻐﺮﻯ ﺍﻟﻤﻮﺯﻭﻧﺔ ﻟﺘﻘﺪﻳﺮ ﺍﻟﺒﻘﺎء ﻟﻠﻨﻮﺫﺝ ﺍﻻﺣﺘﻤﺎﻟﻲ ﺍﻟﻤﺨﺘﻠﻂ )ﺍﻻﺳﻲ –ﻛﺎﻣﺎ ﻣﻦ ﺍﻟﺮﺗﺒﺔ ﺍﻟﺜﺎﻧﻴﺔ(‬

‫ﺍﻟﻣﺳﺗﺧﻠﺹ‬
‫ﻓﻲ ﻫﺬﺍ ﺍﻟﺒﺤﺚ ﺳﻴﺘﻢ ﺗﻘﺪﻳﺮ ﺩﺍﻟﺔ ﺍﻟﺒﻘﺎء ﻟﻠﻨﻤﻮﺫﺝ ﺍﻻﺣﺘﻤﺎﻟﻲ ﺍﻟﻤﺨﺘﻠﻂ ) ﺍﻻﺳﻲ – ﻛﺎﻣﺎ ﻣﻦ ﺍﻟﺮﺗﺒﺔ ﺍﻟﺜﺎﻧﻴﺔ ( ) ‪exponential - gamma of‬‬
‫‪ (the second order‬ﺑﺄﺳﺘﻌﻤﺎﻝ ﻁﺮﻳﻘﺔ ﺍﻟﻤﺮﺑﻌﺎﺕ ﺍﻟﺼﻐﺮﻯ ﺍﻟﻤﻮﺯﻭﻧﺔ )‪ (Weighted Least Squares Method‬ﻟﻌﻴﻨﺔ ﻣﻦ‬
‫ﺍﻟﺒﻴﺎﻧﺎﺕ ﺍﻟﺤﻘﻴﻘﻴﺔ ﺍﻟﺨﺎﺻﺔ ﺑﻤﺮﺿﻰ ﺍﻟﻔﺸﻞ ﺍﻟﻜﻠﻮﻱ ﻓﻲ ﺑﻐﺪﺍﺩ ﻭﻟﻠﻔﺘﺮﺓ ﻣﻦ )‪ ٢٠١٥ / ١ / ١٧‬ﻭﻟﻐﺎﻳﺔ ‪ ( ٢٠١٧ /١٢ / ٢٧‬ﻭﺑﺄﺳﺘﺨﺪﺍﻡ‬
‫ﻣﻌﺎﻳﻴﺮ ﺍﻟﻤﻔﺎﺿﻠﺔ ) ﻣﻌﻴﺎﺭ ﺍﻛﺎﻳﻜﻲ ‪ ،‬ﻣﻌﻴﺎﺭ ﺍﻛﺎﻳﻜﻲ ﺍﻟﻤﺼﺤﺢ ‪ ،‬ﻣﻌﻴﺎﺭ ﺍﻟﻤﻌﻠﻮﻣﺎﺕ ﺍﻟﺒﻴﺰﻱ ( ﺑﻴﻦ ﺍﻟﺘﻮﺯﻳﻌﺎﺕ ﺍﻟﻤﺨﺘﻠﻄﺔ ﻭﺍﻟﺘﻮﺯﻳﻌﺎﺕ‬
‫ﺍﻟﻤﻔﺮﺩﺓ ‪ ،‬ﺗﺘﻀﺢ ﺍﻓﻀﻠﻴﺔ ﺍﻟﻨﻤﻮﺫﺝ ﺍﻟﻤﺨﺘﻠﻂ ﻟﺘﻤﺜﻴﻞ ﺍﻟﺒﻴﺎﻧﺎﺕ ﻋﻠﻰ ﺍﻟﺘﻮﺯﻳﻌﺎﺕ ﺍﻟﻤﻔﺮﺩﺓ ‪ ،‬ﻭﺗﻢ ﺍﻳﻀﺎ ﻓﻲ ﻫﺬﺍ ﺍﻟﺒﺤﺚ ﺍﺳﺘﻌﺮﺍﺽ ﺍﺑﺮﺯ‬
‫ﺍﻟﺪﺭﺍﺳﺎﺕ ﺍﻟﺴﺎﺑﻘﺔ ﻓﻲ ﻫﺬﺍ ﺍﻟﻤﺠﺎﻝ ‪ ،‬ﻭﺍﻳﻀﺎ ﺍﻻﺷﺎﺭﺓ ﺍﻟﻰ ﺑﻌﺾ ﺧﺼﺎﺋﺺ ﺩﺍﻟﺔ ﺍﻟﺒﻘﺎء ﻭﺍﻟﺪﻭﺍﻝ ﺍﻟﻤﺮﺗﺒﻄﺔ ﺑﻬﺎ ﻣﺜﻞ ﺩﺍﻟﺔ ﺍﻟﻜﺜﺎﻓﺔ ﺍﻟﺘﺠﻤﻴﻌﻴﺔ‬
‫ﻟﻠﻔﺸﻞ ﻭﻛﺬﻟﻚ ﺩﺍﻟﺔ ﺍﻟﻔﺸﻞ ﺍﻻﺣﺘﻤﺎﻟﻴﺔ ‪ ،‬ﻭﺍﻳﻀﺎ ﺩﺍﻟﺔ ﺍﻟﻤﺨﺎﻁﺮﺓ ‪.‬‬

‫ﺍﻟﻜﻠﻤﺎﺕ ﺍﻟﻤﻔﺘﺎﺣﻴﺔ ‪ :‬ﻁﺮﻳﻘﺔ ﺍﻟﻤﺮﺑﻌﺎﺕ ﺍﻟﺼﻐﺮﻯ ﺍﻟﻤﻮﺯﻭﻧﺔ ‪ ،‬ﻣﻌﻴﺎﺭ ﺍﻛﺎﻳﻜﻲ ‪ ،‬ﻣﻌﻴﺎﺭ ﺍﻛﺎﻳﻜﻲ ﺍﻟﻤﺼﺤﺢ ‪ ،‬ﻣﻌﻴﺎﺭ ﺍﻟﻤﻌﻠﻮﻣﺎﺕ ﺍﻟﺒﻴﺰﻱ ‪.‬‬

‫‪In this paper, the survival function of the mixed probability distribution (exponential - gamma of the‬‬
‫‪second order) will be estimated (exponential - gamma of the second order) by using the weighted‬‬
‫‪Least Squares method for a sample of real data on patients with kidney failure in Baghdad for the‬‬
‫‪period from (١/١٧/٢٠١٥ to ١٢/٢٧/٢٠١٧) And by using differentiation criteria (Akayki standard,‬‬
‫‪Akayki corrected criterion, Bayesian information standard) between mixed and single distributions,‬‬
‫‪the preference of mixed distribution to represent data over single distributions is evident. The most‬‬
‫‪prominent of previous studies in this field, as well as indicating some properties of the survival‬‬
‫‪function and its associated functions, such as the aggregate failure density function, as well as the‬‬
‫‪.probability failure function, as well as the hazard function‬‬

‫‪Key words: Weighted least squares method, Akayeki criterion, Corrected Akayiki criterion, Bayesian‬‬
‫‪information criterion.‬‬
‫‪ .١‬ﺍﻟﻤﻘﺪﻣﺔ‬

‫ﻋﻤﻞ ﺍﻟﺒﺎﺣﺜﻴﻦ ﻭﺍﻟﻤﺨﺘﺼﻴﻦ ﻓﻲ ﺍﻟﻤﺠﺎﻝ ﺍﻻﺣﺼﺎﺋﻲ ﻓﻲ ﺍﻟﻌﻘﻮﺩ ﺍﻻﺧﻴﺮﺓ ﺑﺸﻜﻞ ﻭﺍﺳﻊ ﻭﻛﺒﻴﺮ ﻋﻠﻰ ﺗﻄﻮﻳﺮ ﺍﻟﺘﻘﻨﻴﺔ ﺍﻻﺣﺼﺎﺋﻴﺔ ﺍﻟﻤﻌﺮﻭﻓﺔ "‬
‫ﺑﺎﻟﺘﻮﺯﻳﻌﺎﺕ ﺍﻻﺣﺘﻤﺎﻟﻴﺔ "‪ "Probability distribution‬ﻭﺍﻻﻧﺘﻘﺎﻝ ﺑﻬﺎ ﻣﻦ ﻣﺮﺣﻠﺔ ﺍﻟﺘﻮﺯﻳﻌﺎﺕ ﺍﻟﻤﻔﺮﺩﺓ ﺍﻟﻰ ﻣﺮﺣﻠﺔ ﺍﻟﺘﻮﺯﻳﻌﺎﺕ ﺍﻟﻤﺮﻛﺒﺔ‬
‫)‪ (Compound distribution‬ﺃﻭ ﺍﻟﺘﻮﺯﻳﻌﺎﺕ ﺍﻟﻤﺨﺘﻠﻄﺔ )‪ (Mixed distributions‬ﺃﻭ ﺍﻟﺘﻮﺯﻳﻌﺎﺕ ﺍﻟﻤﻮﺳﻌﺔ )‪(Extend distribution‬‬
‫ﻭﺫﻟﻚ ﺑﻬﺪﻑ ﺍﻟﺒﺤﺚ ﻋﻦ ﺍﻓﻀﻞ ﺗﻤﺜﻴﻞ ﻟﻠﺒﻴﺎﻧﺎﺕ ﻭﺑﺄﻗﻞ ﺃﺧﻄﺎء ﻭﺍﻟﺴﺒﺐ ﻓﻲ ﺫﻟﻚ ﺃﻧﻪ ﻓﻲ ﺍﻟﻌﺪﻳﺪ ﻣﻦ ﺍﻟﺒﻴﺎﻧﺎﺕ ﺍﻟﺘﻲ ﻳﺘﻢ ﺍﻟﺤﺼﻮﻝ ﻭﺗﻤﺜﻴﻠﻬﺎ ﺑﻴﺎﻧﻴﺎ ً‬
‫ﺗﺄﺧﺬ ﺃﻛﺜﺮ ﻣﻦ ﺷﻜﻞ ﻭﻣﻨﺤﻨﻰ ﻟﺘﻮﺯﻳﻊ ﺍﺣﺘﻤﺎﻟﻲ ﻓﻲ ﺍﻟﻤﺪﺓ ﺍﻟﻤﺪﺭﻭﺳﺔ ﻧﻔﺴﻬﺎ ‪ ،‬ﻭ ِﻣﻦ ﺛ َ ًﻢ ﻳﻜﻮﻥ ﻣﺠﺘﻤﻊ ﺍﻟﺪﺭﺍﺳﺔ ﻣﺘﻜﻮﻧﺎ ً ﻣﻦ ﻣﺰﻳﺞ ﻣﻦ ﺍﻟﻤﺠﺘﻤﻌﺎﺕ‬
‫ﺍﻟﺠﺰﺋﻴﺔ ﻟﺬﺍ ﻓﺎﻥ ﺍﻓﻀﻠﻴﻪ ﻫﺬﻩ ﺍﻻﻧﻮﺍﻉ ﺍﻟﺠﺪﻳﺪﺓ ﻣﻦ ﺍﻟﺘﻮﺯﻳﻌﺎﺕ ﻭﻣﻨﻬﺎ ) ﺍﻟﺘﻮﺯﻳﻌﺎﺕ ﺍﻟﻤﺨﺘﻠﻄﺔ ( ﻭﺩﻗﺘﻬﺎ ﻋﻨﺪ ﺍﺟﺮﺍء ﺣﺴﻦ ﺍﻟﻤﻄﺎﺑﻘﺔ ﺣﺴﻤﺖ‬
‫ﺍﻓﻀﻠﻴﺘﻬﺎ ﻟﺘﻤﺜﻴﻞ ﺑﻴﺎﻧﺎﺕ ﺍﻟﻈﻮﺍﻫﺮ ﺍﻟﻤﺪﺭﻭﺳﺔ ‪.‬‬
‫‪ .٢‬ﺑﻌﺾ ﺍﻟﺪﺭﺍﺳﺎﺕ ﺍﻟﺴﺎﺑﻘﺔ‬

‫ﺻﻔﺤﺔ ‪216‬‬ ‫ﻣﺟﻠﺔ ﺍﻟﺩﻧﺎﻧﻳﺭ ) ﺍﻟﻌﺩ‪(٢٤‬‬

 ‫ﺍﺳﺘﺨﺪﺍﻡ ﻁﺮﻳﻘﺔ ﺍﻟﻠﻤﺮﺑﻌﺎﺕ ﺍﻟﺼﻐﺮﻯ ﺍﻟﻤﻮﺯﻭﻧﺔ ﻟﺘﻘﺪﻳﺮ ﺍﻟﺒﻘﺎء ﻟﻠﻨﻮﺫﺝ ﺍﻻﺣﺘﻤﺎﻟﻲ ﺍﻟﻤﺨﺘﻠﻂ )ﺍﻻﺳﻲ –ﻛﺎﻣﺎ ﻣﻦ ﺍﻟﺮﺗﺒﺔ ﺍﻟﺜﺎﻧﻴﺔ(‬

‫‪ ‬ﻓﻲ ﻋﺎﻡ )‪ (٢٠٠٦‬ﻗﺎﻡ ﺍﻟﺒﺎﺣﺜﺎﻥ )‪[٥] (Ibrahim JG&Chi yy‬ﺑﺪﺭﺍﺳﺔ ﺍﻟﻨﻤﺎﺫﺝ ﺍﻟﻤﺸﺘﺮﻛﺔ ﻟﻠﻤﺘﻐﻴﺮﺍﺕ ﺍﻟﻤﺘﻌﺪﺩﺓ ﺍﻟﻄﻮﻟﻴﺔ‬
‫ﻭﺑﻴﺎﻧﺎﺕ ﺍﻟﺒﻘﺎء ﺍﻟﻤﺘﻌﺪﺩﺓ ﺍﻟﻤﺘﻐﻴﺮﺍﺕ ﻭﺩﺭﺍﺳﺔ ﺍﻻﺭﺗﺒﺎﻁ ﺑﻴﻦ ﺍﻭﻗﺎﺕ ﺍﻟﻔﺸﻞ ﻭﺣﺴﺎﺏ ﺩﺍﻟﺔ ﺍﻟﺒﻘﺎء ﺍﻟﺤﺪﻳﺔ ﻭﺍﻟﺸﺮﻁﻴﺔ ﻣﻦ ﺃﻧﻤﻮﺫﺝ ﺍﻟﺒﻘﺎء‬
‫ﻣﺘﻌﺪﺩ ﺍﻟﻤﺘﻐﻴﺮﺍﺕ ﻭﺗﻄﺒﻴﻘﻪ ﻋﻠﻰ ﻣﺠﻤﻮﻋﺔ ﻣﻦ ﺍﻟﻤﺼﺎﺑﺎﺕ ﺑﺎﻭﺭﺍﻡ ﺍﻟﺜﺪﻱ )‪.(IBCSG‬‬

‫‪ ‬ﻓﻲ ﻋﺎﻡ )‪ (٢٠٠٨‬ﻗﺪﻡ )]‪ (Cox[٦‬ﺑﺤﺜﺎ ﻓﻲ ﺗﻮﺯﻳﻊ ‪ F‬ﺍﻟﻤﻌﻤﻢ )‪ ;(GF‬ﻟﺘﻘﺪﻳﺮ ﺩﺍﻟﺔ ﺍﻟﺒﻘﺎء ﺇﺫ ﻳﺘﻢ ﺗﺼﻨﻴﻒ ﺩﻭﺍﻝ ﺍﻟﻤﺨﺎﻁﺮﺓ ﻟﺘﻮﺯﻳﻊ ‪GF‬‬
‫ﻭﺗﻢ ﻣﻨﺎﻗﺸﺔ ﺗﺮﻛﻴﺐ ‪ GF‬ﻣﻊ ﺑﺪﻳﻞ ﺍﻟﻤﻌﺎﻳﻴﺮ ﻭﺍﻟﺜﻮﺍﺑﺖ ﺑﺎﺳﺘﺨﺪﺍﻡ ﺍﻟﺒﺮﺍﻣﺞ ﺍﻷﺣﺼﺎﺋﻴﺔ ﺍﻟﻤﺸﻬﻮﺭﺓ ﻓﻲ ﻫﺬﺍﺍﻟﻤﺠﺎﻝ ‪ ،‬ﻭﺗﺤﺴﻴﻦ ﻭﺻﻒ‬
‫ﺧﻄﺮ ﺍﻟﻤﻮﺕ ﺑﻌﺪ ﺗﺸﺨﻴﺺ ﺍﻻﻳﺪﺯ ﺍﻟﺴﺮﻳﺮﻱ ﻓﻲ ﺃﺭﺑﻊ ﻓﺘﺮﺍﺕ ﻣﺨﺘﻠﻔﺔ ﻣﻦ ﺍﻟﻌﻼﺝ ﺑﻔﻴﺮﻭﺱ ﻧﻘﺺ ﺍﻟﻤﻨﺎﻋﺔ ﺍﻟﺒﺸﺮﻳﺔ ‪.‬‬
‫‪ ‬ﻓﻲ ﻋﺎﻡ )‪ (٢٠١٠‬ﻗﺎﻡ ﺍﻟﺒﺎﺣﺜﺎﻥ )‪[٨](H. Zakerzadeh and A. Dolati‬ﺑﺪﺭﺍﺳﺔ ﺗﻮﺯﻳﻊ ﻟﻴﻨﺪﻟﻲ ﺍﻟﻌﺎﻡ ﺫﻱ ﺍﻟﻤﻌﻠﻤﺎﺕ ﺍﻟﺜﻼﺙ‬
‫𝛼( ﻭﺗﻢ ﺗﺤﺪﻳﺪ ﺧﺼﺎﺋﺺ ﺩﺍﻟﺔ‬ ‫ﻭﻫﻮ ﺗﻮﺯﻳﻊ ﺧﻠﻴﻂ ﻣﻦ ﺗﻮﺯﻳﻌﻲ ﻛﺎﻣﺎ ﺍﻻﻭﻝ ﺑﺎﻟﻤﻌﻠﻤﺘﻴﻦ ) 𝜃 ‪ (α,‬ﻭﺍﻟﺜﺎﻧﻲ ﺑﺎﻟﻤﻌﻠﻤﺘﻴﻦ ) 𝜃 ‪١,‬‬
‫ﺍﻟﻤﺨﺎﻁﺮﺓ ﻭﺗﻘﺪﻳﺮ ﺍﻟﻤﻌﻠﻤﺎﺕ ﺍﻟﺜﻼﺙ ﺑﻄﺮﻳﻘﺔ ﺍﻻﻣﻜﺎﻥ ﺍﻻﻋﻈﻢ ‪ ،‬ﻭﻛﺬﻟﻚ ﺍﻗﺘﺮﺡ ﺍﻟﺒﺎﺣﺜﺎﻥ ﺧﻮﺍﺭﺯﻣﻴﺔ ﻟﺘﻮﻟﻴﺪ ﻣﺘﻐﻴﺮﺍﺕ ﻋﺸﻮﺍﺋﻴﺔ ﻟﻬﺬﺍ‬
‫ﺍﻟﺘﻮﺯﻳﻊ ‪ ،‬ﻭﺍﻅﻬﺮ ﺍﻟﺘﻄﺒﻴﻖ ﺍﻟﻌﻤﻠﻲ ﺍﻟﺬﻱ ﻗﺎﻡ ﺑﻪ ﺍﻟﺒﺎﺣﺜﺎﻥ ﺍﻥ ﺗﻮﺯﻳﻊ ﻟﻴﻨﺪﻟﻲ ﺍﻟﻌﺎﻡ ﻫﻮ ﻣﻨﺎﻓﺲ ﻗﻮﻱ ﻟﺘﻮﺯﻳﻌﺎﺕ ﺍﻭﻗﺎﺕ ﺍﻟﺤﻴﺎﺓ ﻣﺜﻞ ﻛﺎﻣﺎ‬
‫ﻭﻭﻳﺒﻞ ﻭ)‪.(lognormal‬‬

‫‪ ‬ﻓﻲ ﻋﺎﻡ )‪ (٢٠١٤‬ﻗﺪﻡ ﺍﻟﺒﺎﺣﺜﺎﻥ )‪ [٧](Hiba Z.& MuhammedL.S. Diab‬ﺗﻮﺯﻳﻊ ) ‪Quasi Lindley Geometric‬‬
‫‪ (Distribution‬ﻭﺩﺭﺱ ﺍﻟﺒﺎﺣﺜﺎﻥ ﻣﺠﻤﻮﻋﺔ ﻣﻦ ﺧﺼﺎﺋﺺ ﺍﻟﺘﻮﺯﻳﻊ ﻣﺜﻞ ﺍﻟﻌﺰﻭﻡ ﻭﺩﺍﻟﺔ ﺍﻟﻤﺨﺎﻁﺮﺓ ﻭﺩﺍﻟﺔ ﺍﻟﺒﻘﺎء ﻭﻗﺎﻡ ﺍﻟﺒﺎﺣﺜﺎﻥ ﺑﺘﻄﺒﻴﻖ‬
‫ﻫﺬﺍ ﺍﻟﺘﻮﺯﻳﻊ ﻋﻠﻰ ﺑﻴﺎﻧﺎﺕ ﺣﻘﻴﻘﻴﺔ ﻟﻤﺠﻤﻮﻋﺔ ﻣﻦ ﺍﻟﻤﺼﺎﺑﻴﻦ ﺑﺴﺮﻁﺎﻥ ﺍﻟﻤﺜﺎﻧﺔ ﻭﺍﻅﻬﺮ ﺍﻟﺘﻄﺒﻴﻖ ﺍﻓﻀﻠﻴﺔ ﻫﺬﺍ ﺍﻟﺘﻮﺯﻳﻊ ﻋﻠﻰ ﺗﻮﺯﻳﻊ ﻟﻴﻨﺪﻟﻲ‬
‫ﺍﻟﻬﻨﺪﺳﻲ ﻭﻋﻠﻰ ﺗﻮﺯﻳﻊ ﻟﻴﻨﺪﻟﻲ ‪.‬‬
‫‪ ‬ﻓﻲ ﺍﻟﻌﺎﻡ )‪ (٢٠١٨‬ﻗﺎﻡ ﺍﻟﺒﺎﺣﺚ )ﺣﻤﺰﺓ(]‪ [٢‬ﺑﺘﻘﺪﻳﺮ ﺩﺍﻟﺔ ﺍﻟﺒﻘﺎء ﻟﻤﺮﻳﻀﺎﺕ ﺳﺮﻁﺎﻥ ﺍﻟﺜﺪﻱ ﻓﻲ ﻣﻨﻄﻘﺔ ﺍﻟﻔﺮﺍﺕ ﺍﻻﻭﺳﻂ ﺭﻛﺰﺕ ﻫﺬﻩ‬
‫ﺍﻟﺪﺭﺍﺳﺔ ﻋﻠﻰ ﺗﻘﺪﻳﺮ ﺩﺍﻟﺔ ﺍﻟﺒﻘﺎء ﻭﺍﻟﺪﻭﺍﻝ ﺍﻟﻤﺮﺗﺒﻄﺔ ﺑﻬﺎ )ﺩﺍﻟﺔ ﺍﻟﻜﺜﺎﻓﺔ ﺍﻻﺣﺘﻤﺎﻟﻴﺔ )‪ f(t‬ﻭﺩﺍﻟﺔ ﺍﻟﺘﻮﺯﻳﻊ ﺍﻟﺘﺮﺍﻛﻤﻲ )‪ ،F(t‬ﻭﺩﺍﻟﺔ‬
‫ﺍﻟﺨﻄﻮﺭﺓ )‪ ، h(t‬ﺍﺫ ﺍﺳﺘﺨﺪﻣﺖ ﺛﻼﺙ ﻁﺮﺍﺋﻖ ﻣﻌﻠﻤﻴﻪ ﻟﻤﻌﻠﻤﺎﺕ ﺗﻮﺯﻳﻊ ﻭﻳﺒﻞ ﺫﻭ ﺍﻟﻤﻌﻠﻤﺘﻴﻦ ﻭﺍﻟـــﺪﻭﺍﻝ ﺍﻟﻤـــﺮﺗﺒﻂ ﺑﻬــــﺎ ﻭﻫـــﺬﻩ‬
‫ﺍﻟﻄـــﺮﺍﺋﻖ ﻫـــﻲ ﻁﺮﻳﻘــﺔ ﺍﻻﻣﻜﺎﻥ ﺍﻻﻋﻈﻢ ) ‪ ( MLE‬ﻁﺮﻳﻘﺔ ﺍﻟﻌﺰﻭﻡ )‪ ، (M.O.M‬ﻭﻁﺮﻳﻘﺔ ﺍﻟﻤﺮﺑﻌﺎﺕ ﺍﻟﺼﻐﺮﻯ )‪. (O.L.S‬‬

‫‪ .٣‬ﺍﻟﺠﺎﻧﺐ ﺍﻟﻨﻈﺮﻱ‬

‫‪٢ ٤‬‬
‫‪The Survival Function‬‬ ‫‪ ١.٣‬ﺩﺍﻟﺔ ﺍﻟﺑﻘﺎء‬
‫ﺗﻌﺭﻑ ﺩﺍﻟﺔ ﺍﻟﺑﻘﺎء ﺑﺎﻧﻬﺎ ﺍﺣﺗﻣﺎﻝ ﻋﺩﻡ ﻓﺷﻝ ﺍﻟﻌﻧﺻﺭ ﻓﻲ ﺍﻟﻔﺎﺻﻝ ﺍﻟﺯﻣﻧﻲ )‪ (t,٠‬ﻭﻳﺭﻣﺯ ﻟﻬﺎ ﺑﺎﻟﺭﻣﺯ ‪ S t‬ﻭﻫﻲ ﺩﺍﻟﺔ ﻣﺗﻧﺎﻗﺻﺔ ﻣﻊ‬
‫ﺍﻟﺯﻣﻥ ﻭﺻﻳﻐﺗﻬﺎ ﺍﻟﺭﻳﺎﺿﻳﺔ ﺍﻷﺗﻳﺔ ‪:‬‬

‫‪S t‬‬ ‫‪p T‬‬ ‫𝑡‬ ‫‪f t dt‬‬

‫‪S t‬‬ ‫‪١‬‬ ‫‪p T‬‬ ‫‪t‬‬

‫‪S t‬‬ ‫‪١‬‬ ‫‪F t‬‬ ‫‪… ١‬‬
‫‪ : T‬ﻣﺗﻐﻳﺭ ﻋﺷﻭﺍﺋﻲ ﻳﺭﻣﺯ ﺇﻟﻰ ﺍﻟﻔﺗﺭﺓ ﺍﻟﺯﻣﻧﻳﺔ ﺍﻟﻼﺯﻣﺔ ﻟﺣﺩﻭﺙ ﺍﻟﻔﺷﻝ ‪ ،‬ﺃﻭ ﻫﻭ ﺫﻟﻙ ﺍﻟﻣﺗﻐﻳﺭ ﺍﻟﻌﺷﻭﺍﺋﻲ ﺍﻟﺫﻱ ﻳﺷﻳﺭ ﺇﻟﻰ ﻭﻗﺕ ﺍﻟﺑﻘﺎء‬
‫ﺣﺗﻰ ﺣﺩﻭﺙ ﺍﻟﻣﻭﺕ‬

‫ﺻﻔﺤﺔ ‪217‬‬ ‫ﻣﺟﻠﺔ ﺍﻟﺩﻧﺎﻧﻳﺭ ) ﺍﻟﻌﺩ‪(٢٤‬‬

 ‫ﺍﺳﺘﺨﺪﺍﻡ ﻁﺮﻳﻘﺔ ﺍﻟﻠﻤﺮﺑﻌﺎﺕ ﺍﻟﺼﻐﺮﻯ ﺍﻟﻤﻮﺯﻭﻧﺔ ﻟﺘﻘﺪﻳﺮ ﺍﻟﺒﻘﺎء ﻟﻠﻨﻮﺫﺝ ﺍﻻﺣﺘﻤﺎﻟﻲ ﺍﻟﻤﺨﺘﻠﻂ )ﺍﻻﺳﻲ –ﻛﺎﻣﺎ ﻣﻦ ﺍﻟﺮﺗﺒﺔ ﺍﻟﺜﺎﻧﻴﺔ(‬

‫‪.(t‬‬ ‫‪ :t‬ﻓﻳﻣﺛﻝ ﺯﻣﻥ ﺍﻟﺑﻘﺎء ﺍﻟﺫﻱ ﻳﻛﻭﻥ ﺃﻛﺑﺭ ﺃﻭ ﻳﺳﺎﻭﻱ ﺻﻔﺭ )‪٠‬‬

‫ﺎو واﺣ ‪ ،‬و ﻠ ﺎ ﯾ داد ﻋ‬ ‫ﺎب ﻋﻠﻰ ﻗ اﻟ ﺎة ﻓﻲ اﻟ ﻣ )‪(٠‬‬ ‫أن اﺣ ﺎل ﻘﺎء اﻟ‬ ‫‪S ٠‬أ‬ ‫ﻏﺎﻟ ﺎ ﻣﺎﻧﻔ ض ان ‪١‬‬

‫اﻟ ﺎﺋ اﻟ ﻲ )اﻻﻧ ﺎن ﻣ ﻼ( ﻘ ب ﻣ اﻟ ﻔ ‪ ،‬ﺍﻱ ﻋﻧﺩﻣﺎ )‪ (t ٠‬ﻓﺎﻥ ﺩﺍﻟﺔ ﺍﻟﺑﻘﺎء ﺳﺗﻛﻭﻥ ﻣﺳﺎﻭﻳﺔ ﻟﻠﻭﺍﺣﺩ ﻭﻳﻌﻧﻲ ﻫﺫﺍ ﺍﻥ ﺍﺣﺗﻣﺎﻝ‬

‫ﺑﻘﺎء ﺍﻟﻔﺭﺩ ﻋﻠﻰ ﻗﻳﺩ ﺍﻟﺣﻳﺎﺓ ﻋﻧﺩ ﺍﻟﻭﻗﺕ )‪ (t ٠‬ﻣﺳﺎﻭﻳﺎ ﻟﻠﻭﺍﺣﺩ ‪.‬‬

‫اﻻ‬ ‫ة ﻣ اﻟ ﺎﻧ‬ ‫ﺎ اﺳﻠﻔ ﺎ داﻟﺔ ﻣ ﺎﻗ ﺔ ﻣﻊ اﻟ ﻣ و ﻟ ﻣ‬ ‫داﻟﺔ اﻟ ﻘﺎء‬ ‫وﻣ أﺑ ز ﺧ ﺎﺋ‬

‫‪S u‬‬ ‫‪s t if u‬‬ ‫𝑡‬

‫)‪(١)(٥‬‬
‫‪Cumulative Density Function‬‬ ‫‪ ٢.٣‬ﺩﺍﻟﺔ ﺍﻟﻜﺜﺎﻓﺔ ﺍﻟﺘﺠﻤﻴﻌﻴﺔ ﻟﻠﻔﺸﻞ‬

‫ﻫﻲ ﺩﺍﻟﺔ ﺍﺣﺗﻣﺎﻟﻳﺔ ﻣﻭﺕ ﺍﻟﻛﺎﺋﻥ ﻗﺑﻝ ﺍﻟﻭﻗﺕ ‪ t‬ﻭﻳﺭﻣﺯ ﻟﻬﺎ ﺑﺎﻟﺭﻣﺯ ‪ F t‬ﻭﺗﺳﻣﻰ ﺍﻳﺿﺎ ﺑﺩﺍﻟﺔ ﺗﻭﺯﻳﻊ ﻭﻗﺕ ﺍﻟﺣﻳﺎﺓ ﻭﻫﻲ )ﻣﻛﻣﻠﺔ ﺩﺍﻟﺔ‬
‫ﺍﻟﺑﻘﺎء( ﻭﺍﻟﺗﻌﺑﻳﺭ ﺍﻟﺭﻳﺎﺿﻲ ﻟﻬﺎ ﻫﻭ ‪:‬‬

‫‪F t‬‬ ‫‪p T‬‬ ‫‪t‬‬

‫‪F t‬‬ ‫‪f u du‬‬ ‫‪… ٢‬‬

‫‪٠‬‬

‫‪F t‬‬ ‫‪١‬‬ ‫‪S t‬‬ ‫𝟑 …‬

‫)‪( Failure Density function‬‬ ‫)‪(١) (٥‬‬

‫‪ ٣.٣‬ﺩﺍﻟﺔ ﺍﻟﻛﺛﺎﻓﺔ ﺍﻻﺣﺗﻣﺎﻟﻳﺔ ﻟﻠﻔﺷﻝ‬

‫‪t٢‬‬ ‫‪t١‬‬ ‫‪ (t,t‬ﻭﺑﻐﺽ ﺍﻟﻧﻅﺭﻋﻥ ﺻﻐﺭ ‪ ∆t‬ﺣﻳﺙ ﺍﻥ ‪∆t‬‬ ‫ﻭﻫﻲ ﺍﺣﺗﻣﺎﻝ ﻓﺷﻝ ﺍﻟﻣﻔﺭﺩﺓ )ﻣﻭﺕ ﺍﻻﻧﺳﺎﻥ( ﺧﻼﻝ ﺍﻟﻣﺩﺓ )‪∆t‬‬
‫ﻭﻳﺭﻣﺯ ﻟﻬﺎ ﺑﺎﻟﺭﻣﺯ ‪ f t‬ﻭﺍﻟﺗﻌﺑﻳﺭ ﺍﻟﺭﻳﺎﺿﻲ ﻟﻬﺎ ﻫﻭ ‪:‬‬

‫‪Pr t‬‬ ‫𝑡 𝑇‬ ‫𝑡∆‬

‫‪f t‬‬ ‫‪lim‬‬ ‫‪,t‬‬ ‫‪٠… ٤‬‬
‫‪∆ →٠‬‬ ‫‪∆t‬‬

‫‪∆t‬‬ ‫‪t‬‬ ‫‪t‬‬ ‫‪١‬‬ ‫ﻭﺍﻥ ‪ : ∆t‬ﻫﻭ ﺍﻟﺗﻐﻳﺭ ﻓﻲ ﻗﻳﻣﺔ ﺍﻟﻣﺗﻐﻳﺭ ﺍﻟﻌﺷﻭﺍﺋﻲ‪ T‬ﺍﻱ ﺑﻣﻌﻧﻰ‬

‫‪١٠ ١٢‬‬
‫ﺗﺳﻣﻰ ﺍﻳﺿﺎ ﻣﻌﺩﻝ ﺍﻟﻔﺷﻝ ﻭﻫﻲ‬ ‫‪Hazard Function‬‬ ‫‪ ٤.٣‬ﺩﺍﻟﺔ ﺍﻟﻣﺧﺎﻁﺭﺓ‬
‫ﺇﺣﺗﻣﺎﻝ ﻓﺷﻝ ﺍﻟﻣﻔﺭﺩﺓ ﺃﻭﺍﻟﻧﻅﺎﻡ ﺧﻼﻝ ﺍﻟﻔﺗﺭﺓ ﺍﻟﺯﻣﻧﻳﺔ )‪ (t,t ∆t‬ﻋﻠﻣﺎ ً ﺃﻥ ﺍﻟﻣﻔﺭﺩﺓ ﺃﻭ ﺍﻟﻧﻅﺎﻡ ﻳﻌﻣﻝ )ﻟﻡ ﻳﻔﺷﻝ( ﺣﺗﻰ ﺍﻟﺯﻣﻥ ‪ t‬ﺍﻣﺎ ﺑﺎﻟﻧﺳﺑﺔ‬
‫ﻟﻠﻛﺎﺋﻥ ﺍﻟﺣﻲ ﻓﻬﻲ ﺗﻣﺛﻝ ﻣﻌﺩﻝ ﺍﻟﻭﻓﻳﺎﺕ ﺍﻟﻠﺣﻅﻳﺔ ﻟﻠﻔﺭﺩ ﺍﻟﺫﻱ ﻳﻧﺟﻭ ﻣﻥ ﺍﻟﺯﻣﻥ ‪ t‬ﺍﻭ ﺍﺣﺗﻣﺎﻝ ﺍﻥ ﺗﻧﺗﻬﻲ ﺍﻟﺣﺎﻟﺔ ﻓﻲ ﺍﻟﻭﻗﺕ ‪ t‬ﻭﻳﺭﻣﺯ ﻟﻬﺎ‬
‫ﺑﺎﻟﺭﻣﺯ ‪ h t‬ﺃﻱ ﺃﻥ‪:‬‬

‫‪Pr t‬‬ ‫𝑇‬ ‫𝑇 ⃥ 𝑡∆ 𝑡‬ ‫𝑡‬

‫‪h t‬‬ ‫‪… ٥‬‬
‫‪∆t‬‬
‫ﺻﻔﺤﺔ ‪218‬‬ ‫ﻣﺟﻠﺔ ﺍﻟﺩﻧﺎﻧﻳﺭ ) ﺍﻟﻌﺩ‪(٢٤‬‬
 ‫ﺍﻟﻤﺨﺘﻠﻂ )ﺍﻻﺳﻲ –ﻛﺎﻣﺎ ﻣﻣﻦ ﺍﻟﺮﺗﺒﺔ ﺍﻟﺜﺎﻧﻴﺔ(‬
‫ﻂ‬ ‫ﻁﺮﻳﻘﺔ ﺍﻟﻠﻤﺮﺑﻌﺎﺕ ﺍﻟﺼﻐﺮﺮﻯ ﺍﻟﻤﻮﺯﻭﻧﺔ ﻟﺘﻘﺪﺪﻳﺮ ﺍﻟﺒﻘﺎء ﻟﻠﻨﻮﺫﺝ ﺍﻻﺣﺘﻤﺎﻟﻲ‬
‫ﺔ‬ ‫ﺍﺳﺘﺨﺪﺍﻡ‬

‫‪ h t‬ﻭﺑﺎﻟﺷﻛﻝ ﺍﻵﺗﻲ‬
‫ﻲ‪:‬‬ ‫ﻧﺣﺻﻝ ﻋﻠﻰ ﺩﺍﻟﺔ ﺍﻟﻣﻣﺧﺎﻁﺭﺓ‬
‫ﻝ‬ ‫ﻭﻋﻧﺩﻣﺎ ‪∆t→ ٠‬‬

‫‪Pr t , t‬‬ ‫‪∆ t \t‬‬

‫‪h t‬‬ ‫‪lim‬‬ ‫‪… ٦‬‬
‫‪∆ →٠‬‬ ‫‪∆t‬‬

‫ﻫﻭ ﺍﻟﺩﺍﻟﺔ ﺍﻟﻣﺷﺗﺭﻛﻛﺔ ‪ joint‬ﻣﻘﻘﺳﻭﻣﻪ ﻋﻠﻰ ﺍﻟﺩﺍﻟﺔ ﺍﻟﺣﺩﻳﺔ ‪nal‬‬

‫‪Marjin‬‬ ‫ﻭﺍﻻﺣﺗﻣﺎﺎﻝ ﺍﻟﺷﺭﻁﻲ ﻭ‬

‫𝑡∆ 𝑡 𝑇 ‪p t‬‬
‫‪pr‬‬
‫‪h t‬‬ ‫‪lim‬‬ ‫‪… ٧‬‬
‫‪∆ →٠‬‬ ‫𝑡 ‪∆t pr T‬‬

‫‪١‬‬ ‫‪pr t‬‬ ‫𝑡 𝑇‬ ‫𝑡𝑡∆‬

‫‪h t‬‬ ‫‪lim‬‬ ‫‪… ٨‬‬
‫‪S t ∆ →٠‬‬ ‫‪∆t‬‬

‫‪dF t‬‬ ‫‪١‬‬ ‫‪f t‬‬

‫‪h t‬‬ ‫‪.‬‬ ‫‪… ٩‬‬
‫‪dt‬‬ ‫‪S t‬‬ ‫‪S t‬‬

‫‪ Severity function‬ﺗﺗﻧﺎﺎﺳﺏ ﻋﻛﺳﻳﺎ ً ﻣﻊ ‪ survvival funcctionS t‬ﻭﻁﺭﺩﻳﺎ ً ﻣﻊ ‪density‬‬

‫‪Probability d‬‬ ‫‪nh‬‬ ‫ﺇﻥ ‪t‬‬
‫‪ fun‬ﻟﺫﻟﻙ ﻥ‬
‫ﻓﺎﻥ ﺍﻟﻌﻼﻗﺔ ﺍﻟﺗﻲ ﺗﺭﺑﻁ ﻫﺫﻩ ﺍﻟﻣﻔﻔﺭﺩﺍﺕ ﺍﻟﺛﻼﺙ ﻭﺍﻟﻣﺗﻣﺛﻠﺔ ﺑﺎﻟﻣﻌﻌﺎﺩﻟﺔ )‪ (٩‬ﻥ‬
‫ﻳﻣﻛﻥ ﺍﺳﺗﻐﻼﻟﻬﺎ ﻭﻭﺗﻁﺑﻳﻘﻬﺎ ﻓﻲ ﺍﻟﻭﻭﺍﻗﻊ‪،‬‬ ‫‪nctionf t‬‬
‫ﻋﻠﻰ ﺇﺛﻧﻳﻥ ﻣﻥ ﻫﻫﺫﻩ ﺍﻟﺩﻭﺍﻝ ﻳﻣﻛﻧﻧﻧﺎ ﻣﻥ ﻣﻌﺭﻓﺔ ﺍﻟﺩﺍﻟﺔ ﺍﻟﺛﺎﻟﺛﺔ‪.‬‬
‫ﻭﺍﻟﺣﺻﻭﻝ ﻋ‬

‫ﺣﻭﺭ ﺍﻻﻓﻘﻲ ﺍﻟﻭﻭﻗﺕ ‪ t‬ﻭﺍﻟﻣﺣ‬

‫ﺣﻭﺭ ﺍﻟﻌﻣﻭﺩﻱ ﻳﺗﻣﺛﻝ ﺑﻘﻳﻣﺔ ﺩﺍﻟﺔ‬ ‫ﺣﻳﺙ ﺍﻥ ﺍﻟﻣﺣ‬
‫ﻭ ﺍﻟﺷﻛﻝ ﺭﺭﻗﻡ )‪ (١‬ﻳﺑﻳﻥ ﻣﻧﺣﻧﻰ ﺩﺍﻟﺔ ﺍﻟﻣﺧﺎﻁﺭﺓ‪ ،‬ﺑﺣ‬
‫ﻭﺍﻟﺗﻲ ﻛﻠﻣﺎ ﺗﻘﺩﻡ ﻋﻣﺭﺭﺍﻟﻛﺎﺋﻥ ﺍﻟﺣﻲ ﺯﺯﺍﺩﺕ‬
‫ﻲ‬ ‫ﺧﺎﻁﺭﺓ ‪h t‬ﻭﺍﻟﻟﺯﻣﻥ ‪t‬‬
‫ﺍﻟﺗﻧﺎﺳﺏ ﺍﻟﻁﺭﺩﻱ ﺑﻳﻥ ﻗﻳﻣﺔ ﺩﺍﻟﺔ ﺍﻟﻣﺧ‬
‫ﺏ‬ ‫ﺍﻟﻣﺧﺎﻁﺭﺓ ‪h t‬ﻭﻧﻼﺣﻅ ﻥ‬
‫ﻣﻥ ﺍﻟﺷﻛﻝ‬
‫ﺍﻟﺧﻁﻭﺭﺓ ‪.‬‬

‫‪Time‬‬

‫ﺓ‬
‫ﻣﺧﺎﻁﺭﺓ‬ ‫ﺷﻛﻝ )‪ (١‬ﻳﻭﺿﺢ ﻣﻧﺣﻧﻰ ﺩﺍﻟﺔ ﺍﻟ‬

‫ﺻﻔﺤﺔ ‪219‬‬ ‫ﻣﺟﻠﺔ ﺍﻟﺩﻧﺎﻧﻳﺭ ) ﺍﻟﻌﺩ‪(٢٤‬‬

 ‫ﺍﻟﻤﺨﺘﻠﻂ )ﺍﻻﺳﻲ –ﻛﺎﻣﺎ ﻣﻣﻦ ﺍﻟﺮﺗﺒﺔ ﺍﻟﺜﺎﻧﻴﺔ(‬
‫ﻂ‬ ‫ﻁﺮﻳﻘﺔ ﺍﻟﻠﻤﺮﺑﻌﺎﺕ ﺍﻟﺼﻐﺮﺮﻯ ﺍﻟﻤﻮﺯﻭﻧﺔ ﻟﺘﻘﺪﺪﻳﺮ ﺍﻟﺒﻘﺎء ﻟﻠﻨﻮﺫﺝ ﺍﻻﺣﺘﻤﺎﻟﻲ‬
‫ﺔ‬ ‫ﺍﺳﺘﺨﺪﺍﻡ‬

‫‪ ٥.٣‬ﺍﻟﻧﻣﻭﺫﺫﺝ ﺍﻻﺣﺗﻣﺎﻟﻲ ﺍﻟﻣﺧﺗﻠﻁ ) ﻲ‬

‫ﺍﻻﺳﻲ –ﻛﺎﻣﺎ ﻣﻥ ﺍﻟﺭﺭﺗﺑﺔ ﺍﻟﺛﺎﻧﻳﺔ(‬

‫‪Mixed‬‬
‫‪M‬‬ ‫‪probability distrribution exponential - gamma of‬‬
‫‪o the secon‬‬
‫‪nd order‬‬
‫ﻻﺗﺄﺧﺬ‬
‫ﺸﺎﻫﺪﺍﺕ( ﻏﻴﺮ ﺍﺍﻟﻤﺘﺠﺎﻧﺴﺔ ‪ ،‬ﻭﻻ‬ ‫ﺍﻛﺛﺭ ﻭﻳﺳﺗﺧﺩﻡ ﻥ‬
‫ﺗﻜﻮﻥ ﺍﻟﺒﻴﺎﻧﺎﺕ )ﺍﻟﻤﺸ‬ ‫ﻻﺣﺗﻣﺎﻟﻲ ﺍﻟﻣﺧﺗﻠﻠﻁ ﻫﻭ ﺧﻠﻳﻁ ﻧﺎﺗﺗﺞ ﻋﻥ ﺩﻣﺞ ﺗﻭﻭﺯﻳﻌﻳﻥ ﺍﻭ ﺭ‬
‫ﺍﻟﻧﻣﻭﺫﺝ ﺍﻻ‬
‫ﺍﻟﺘﻮﺯﻳﻊ ﺫﺍﺗﻪ ﺑﻤﻌﻠﻤﺎﺕ ﻣﺨﺘﻠﻔﺔ ﺃﻭ ﻗﺪ ﺗﺗﻜﻮﻥ‬
‫ﻊ‬ ‫ﺣﻴﺚ ﺍﻥ ﻛﻞ ﺟﺰﺰء ﻣﻨﻬﺎ ﻳﺘﺒﻊ ﺗﻮﻮﺯﻳﻌﺎ ً ﻣﻌﻴﻨﺎ ‪ ،‬ﺪ‬
‫ﻗﺪ ﻳﻜﻮﻥ‬ ‫ﺍﻟﺘﻮﺯﻳﻌﺎﺕ ﺍﻻﺣﺘﻤﺎﻟﻴﺔ ﺣ‬
‫ﺕ‬ ‫ﻭﺍﺣﺪﺍ ً ﻣﻦ‬
‫ﺷﻜﻼً ﺪ‬
‫ﺴﻠﻮﻙ‬
‫ﺠﺘﻤﻌﺎﺕ ﺍﻟﺠﺰﺋﻴﻴﺔ ﺍﻷﺧﺮﻯ ‪ ،‬ﻭﻭﺑﺎﻟﺘﺎﻟﻲ ﻓﺎﻥ ﺍﻟﺴ‬
‫ﻛﺜﺎﻓﺔ ﺍﺣﺘﻤﺎﻟﻴﺔ ﻗﺪ ﺗﺗﺨﺘﻠﻒ ﻋﻦ ﺍﻟﻤﺠ‬
‫ﻣﺠﺘﻤﻊ ﻟﺪﻳﻪ ﺩﺍﻟﺔ ﺔ‬
‫ﻊ‬ ‫ﻜﻮﻥ ﻛﻞ‬
‫ﺗﻮﺯﻳﻌﺎﺕ ﻣﻣﺨﺘﻠﻔﺔ ‪ ،‬ﺍﺫ ﻳﻜﻮ‬
‫‪Mix‬‬
‫ﻄﻠﻖ ﻋﻠﻰ ﻫﺬﺍ ﺍﻟﻨﻟﻨﻮﻉ ﺑﺎﻟﺘﻮﺯﻳﻊ ﺍﻟﻟﻤﺨﺘﻠﻂ ) ‪xture‬‬
‫ﻟﻠﻤﺠﺘﻤﻌﺎﺕ ﺍﻟﺠﺰﺋﻴﺔ ‪ ،‬ﻭﻳﻄ‬
‫ﺕ‬ ‫ﺍﻻﺣﺘﻤﺎﻟﻲ ﻟﻠﻤﺠﺘﻤﻊ ﻛﻜﻞ ﻳﻳﻤﺜﻞ ﻣﺰﻳﺞ ﻣﻦ ﺍﻟﺴﻠﻮﻙ ﺍﻻﺣﺘﻤﺎﺎﻟﻲ‬
‫‪ . (distrribution‬ﻭﺍﻟﺗﻭﺯﺯﻳﻊ ﺍﻟﻣﺧﺗﻠﻁ ﺍﻟﻟﻣﺩﺭﻭﺱ ﻫﻧﺎ ﻧﺎﺎﺗﺞ ﻋﻥ ﻣﺗﻐﻳﺭﻳﻳﻥ ﻋﺷﻭﺍﺋﻳﻳﻥ ﺍﺍﺣﺩﻫﻣﺎ ﻳﺗﺑﻊ ﺍﻟﺗﺗﻭﺯﻳﻊ ﺍﻻﺳﻲ ﺑﺑﻣﻌﻠﻣﺔ )‪ (λ‬ﻭﺍﻻﺧﺭ‬
‫ﻳﺗﻭﺯﻉ ﻛﺎﻣﻣﺎ ﺑﻣﻌﻠﻣﺔ ﺷﻛﻝ ﻗﻳﻣﺗﻬﺎ ﺗﺳﺎﻭﻱ )‪ (٣‬ﻭﻣﻌﻠﻣﺔ ﻗﻳﻳﺎﺱ )‪. (λ‬‬
‫‪ -‬ﺩﺍﻟﺔ ﺍﻟﻛﺗﺗﻠﺔ ﺍﻻﺣﺗﻣﺎﻟﻳﺔ ﻟﻟﻠﻧﻣﻭﺫﺝ ﺍﻻﺣﺗﻣﻣﺎﻟﻲ ﺍﻟﻣﺧﺗﻠﻁ ‪:‬‬

‫𝛽‬ ‫‪١‬‬ ‫‪𝜆٣ ٢‬‬

‫‪f‬‬ ‫‪t; λ, β‬‬ ‫‪λe‬‬ ‫𝑒 𝑡‬ ‫𝟎𝟏 …‬
‫𝛽‬ ‫‪١‬‬ ‫𝛽‬ ‫‪١٢‬‬

‫‪λ‬‬ ‫‪𝜆٢ ٢‬‬

‫‪f‬‬ ‫‪t; λ, β‬‬ ‫‪e‬‬ ‫𝛽‬ ‫𝑡‬ ‫𝟏𝟏 …‬
‫𝛽‬ ‫‪١‬‬ ‫‪٢‬‬

‫‪(pd‬ﻟﻠﻧﻣﻭﺫﺝ ﺍﻻﻻﺣﺗﻣﺎﻟﻲ ﺍﻟﻣﺧﺗﻠﻠﻁ ﻟﻌﺩﺩ ﻣﻥ ﺍﻟﻣﻣﻌﻠﻣﺎﺕ‬

‫ﺷﻛﻝ )‪ (٢‬ﻣﻣﻧﺣﻧﻰ ﺩﺍﻟﺔ )‪df‬‬

‫ﻭﺍﻟﺩﺍﻟﺔ ﺍﻟﺗﺗﺟﻣﻳﻌﻳﺔ ﻟﻬﺫﺍ ﺍﻟﻟﺗﻭﺯﻳﻊ ﻛﺎﻷﺗﻲ ‪:‬‬

‫ﺻﻔﺤﺔ ‪220‬‬ ‫ﻣﺟﻠﺔ ﺍﻟﺩﻧﺎﻧﻳﺭ ) ﺍﻟﻌﺩ‪(٢٤‬‬

 ‫ﺍﻟﻤﺨﺘﻠﻂ )ﺍﻻﺳﻲ –ﻛﺎﻣﺎ ﻣﻣﻦ ﺍﻟﺮﺗﺒﺔ ﺍﻟﺜﺎﻧﻴﺔ(‬
‫ﻂ‬ ‫ﻁﺮﻳﻘﺔ ﺍﻟﻠﻤﺮﺑﻌﺎﺕ ﺍﻟﺼﻐﺮﺮﻯ ﺍﻟﻤﻮﺯﻭﻧﺔ ﻟﺘﻘﺪﺪﻳﺮ ﺍﻟﺒﻘﺎء ﻟﻠﻨﻮﺫﺝ ﺍﻻﺣﺘﻤﺎﻟﻲ‬
‫ﺔ‬ ‫ﺍﺳﺘﺨﺪﺍﻡ‬

‫𝛽‬ ‫‪١‬‬ ‫‪𝜆٢ 𝑡 ٢‬‬

‫‪F‬‬ ‫‪t; λ, β‬‬ ‫‪١‬‬ ‫‪e‬‬ ‫‪١‬‬ ‫‪e‬‬ ‫‪١‬‬ ‫𝑡𝜆‬ ‫𝟐𝟏 …‬
‫𝛽‬ ‫‪١‬‬ ‫𝛽‬ ‫‪١‬‬ ‫‪٢‬‬

‫ﻣﻌﻠﻣﺎﺕ‬
‫ﺍﻟﻣﺧﺗﻠﻁ ﻟﻌﺩﺩ ﻣﻥ ﺍﻟﻣﻌ‬
‫ﻠﻁ‬ ‫‪(cd‬ﻟﻠﻧﻣﻭﺫﺝ ﺍﻻﻻﺣﺗﻣﺎﻟﻲ‬
‫ﺷﻛﻝ )‪ (٣‬ﻣﻣﻧﺣﻧﻰ ﺩﺍﻟﺔ )‪df‬‬
‫ﺣﺗﻣﺎﻟﻲ ﺍﻟﻣﺧﺗﻠﻁ ‪:‬‬
‫ﺍﻟﺑﻘﺎء ﻟﻠﻧﻣﻭﺫﺝ ﺍﻻﺣﺗ‬
‫ء‬ ‫‪ -‬ﺩﺍﻟﺔ‬

‫‪S‬‬ ‫‪t; λ, β‬‬ ‫‪١‬‬ ‫‪F‬‬ ‫‪t; λ, β‬‬

‫𝛽‬ ‫‪١‬‬ ‫‪𝜆٢ 𝑡 ٢‬‬

‫‪S‬‬ ‫‪t; λ, β‬‬ ‫‪e‬‬ ‫‪e‬‬ ‫‪١‬‬ ‫𝑥𝜆‬
‫𝜆‬ ‫𝟑𝟏 …‬
‫𝛽‬ ‫‪١‬‬ ‫𝛽‬ ‫‪١‬‬ ‫‪٢‬‬

‫ﺍﻻﺣﺗﻣﺎﻟﻲ ﺍﻟﻣﺧﺗﻠﻁ ﻟﻌﺩﺩ ﻣﻥ ﺍﻟﻣﻌﻠﻣﺎﺕ‬

‫ﻲ‬ ‫ﺷﻛﻝ )‪ (٤‬ﻣﻧﺣﻧﻰ )‪(Survivaal Function‬ﻟﻠﻧﻣﻭﻭﺫﺝ‬
‫ﻝ‬

‫ﺻﻔﺤﺔ ‪221‬‬ ‫ﻣﺟﻠﺔ ﺍﻟﺩﻧﺎﻧﻳﺭ ) ﺍﻟﻌﺩ‪(٢٤‬‬

 ‫ﺍﺳﺘﺨﺪﺍﻡ ﻁﺮﻳﻘﺔ ﺍﻟﻠﻤﺮﺑﻌﺎﺕ ﺍﻟﺼﻐﺮﻯ ﺍﻟﻤﻮﺯﻭﻧﺔ ﻟﺘﻘﺪﻳﺮ ﺍﻟﺒﻘﺎء ﻟﻠﻨﻮﺫﺝ ﺍﻻﺣﺘﻤﺎﻟﻲ ﺍﻟﻤﺨﺘﻠﻂ )ﺍﻻﺳﻲ –ﻛﺎﻣﺎ ﻣﻦ ﺍﻟﺮﺗﺒﺔ ﺍﻟﺜﺎﻧﻴﺔ(‬

‫‪٩,١١‬‬
‫)‪(Weighted Least Squares‬‬ ‫‪.٦.٣‬ﻁﺭﻳﻘﺔ ﺍﻟﻣﺭﺑﻌﺎﺕ ﺍﻟﺻﻐﺭﻯ ﺍﻟﻣﻭﺯﻭﻧﺔ‬

‫ﻭﻫﻲ ﻁﺭﻳﻘﺔ ﺃﺳﺎﺳﻳﺔ ﻟﻠﺗﻘﺩﻳﺭ ﻭﺗﺗﺿﻣﻳﻥ ﻋﺎﻣﻝ ﺍﻟﻭﺯﻥ ) ‪ (W‬ﻭﻳﻣﻳﺯ ﻫﺫﻩ ﺍﻟﻁﺭﻳﻘﺔ ﻋﻥ ﻁﺭﻳﻘﺔﺍﻟﻣﺭﺑﻌﺎﺕ ﺍﻟﺻﻐﺭﻯ ﺍﻟﻌﺎﺩﻳﺔ ﺑﺎﻷﺳﺗﻧﺎﺩ ﻋﻠﻰ ﻓﻛﺭﺓ‬
‫‪ x‬ﺗﻣﺛﻝ ﺍﻻﺣﺻﺎءﺍﺕ ﺍﻟﻣﺭﺗﺑﺔ ﻟﻘﻳﻡ ﺍﻟﻌﻳﻧﺔ ‪،‬‬ ‫‪.‬‬ ‫‪i‬‬ ‫ﺗﻘﻠﻳﻝ ﻣﺟﻣﻭﻉ ﻣﺭﺑﻌﺎﺕ ﺍﻷﺧﻁﺎء ﻭﺷﻛﻠﻬﺎ ﻗﺩﺭ ﺍﻹﻣﻛﺎﻥ‪ ،‬ﻭﺑﺎﻓﺗﺭﺍﺽ ﺍﻥ ‪١,٢ … . , n‬‬
‫‪ G x‬ﻭﺍﻥ ﺍﻟﺗﻭﻗﻊ ﻭﺍﻟﺗﺑﺎﻳﻥ ﻟﻬﺫﺍ ﺍﻟﺗﻭﺯﻳﻊ ﻫﻣﺎ ﻋﻠﻰ ﺍﻟﺗﻭﺍﻟﻲ‪:‬‬ ‫ﻓﺎﻥ ﺍﻟﻁﺭﻳﻘﺔ ﺍﻟﻣﻘﺗﺭﺣﺔ ﻫﻲ ﺍﺳﺗﺧﺩﺍﻡ ﺗﻭﺯﻳﻊ‬

‫‪E G x‬‬
‫‪١‬‬

‫‪i n i ١‬‬
‫‪V G x‬‬
‫‪n ١ ٢ n ٢‬‬

‫ﻭﺑﺎﺳﺗﺧﺩﺍﻡ ﺍﻟﺗﻭﻗﻊ ﻭﺍﻟﺗﺑﺎﻳﻥ ﻳﻣﻛﻥ ﺍﻟﺣﺻﻭﻝ ﻋﻠﻰ ﺍﻟﻣﻘﺩﺭﺍﺕ ﺑﻁﺭﻳﻘﺔ ﺍﻟﻣﺭﺑﻌﺎﺕ ﺍﻟﺻﻐﺭﻯ ﺍﻟﻣﻭﺯﻭﻧﺔ ‪ ،‬ﺑﺗﻁﺑﻳﻖ ﺑﻬﺫﻩ ﺍﻟﻁﺭﻳﻘﺔ ﻭﺍﻟﺗﻲ ﺻﻳﻐﺗﻬﺎ‪:‬‬
‫‪٢‬‬
‫‪i‬‬
‫‪K‬‬ ‫‪W F x‬‬ ‫𝟒𝟏 …‬
‫‪n‬‬ ‫‪١‬‬
‫‪١‬‬

‫ﺣﻳﺙ ﺍﻥ ‪ W‬ﺗﺳﺎﻭﻱ ‪:‬‬

‫‪١‬‬ ‫‪n ١ ٢ n ٢‬‬

‫‪W‬‬
‫‪V F x‬‬ ‫‪i n i ١‬‬

‫‪ λ‬ﻳﻣﻛﻥ ﺍﻟﺣﺻﻭﻝ‬ ‫‪β‬‬ ‫‪,‬‬ ‫ﻭﺑﺎﻟﻧﺳﺑﺔ ﻟﻠﻧﻣﻭﺫﺝ ﺍﻻﺣﺗﻣﺎﻟﻲ ﺍﻟﻣﺧﺗﻠﻁ )‪ (EG‬ﻓﺎﻥ ﻣﻘﺩﺭﺍﺕ ﺍﻟﻣﺭﺑﻌﺎﺕ ﺍﻟﺻﻐﺭﻯ ﺍﻟﻣﻭﺯﻭﻧﺔ ﻭﻫﻣـــــــــــــــﺎ‬
‫ﻋﻠﻳﻬﻣﺎ ﻋﻥ ﻁﺭﻳﻖ ﺗﻘﻠﻳﻝ ﺍﻟﻣﻘﺩﺍﺭ ‪ K‬ﻭﺑﺎﺳﺗﺧﺩﺍﻡ ﺍﻟﺩﺍﻟﺔ ﺍﻟﺗﺭﺍﻛﻣﻳﺔ ‪:‬‬

‫‪٢‬‬
‫‪١‬‬ ‫‪e‬‬ ‫‪𝜆٢ 𝑥 ٢‬‬ ‫‪i‬‬
‫‪K‬‬ ‫‪W‬‬ ‫‪β‬‬ ‫‪١‬‬ ‫‪λx‬‬ ‫‪… ١٥‬‬
‫𝛽‬ ‫‪١‬‬ ‫‪٢‬‬ ‫‪n‬‬ ‫‪١‬‬
‫‪١‬‬

‫ﻭﺑﺄﺷﺗﻘﺎﻕ ﺍﻟﻣﻌﺎﺩﻟﺔ )‪ (١٤‬ﻟﻠﻣﻌﻠﻣﺔ )‪ (β‬ﻭﺍﻟﻘﺳﻣﺔ ﻋﻠﻰ ‪ ٢‬ﻧﺣﺻﻝ ﻋﻠﻰ ‪:‬‬

‫‪dK‬‬ ‫‪١‬‬ ‫‪e‬‬ ‫‪𝜆٢ 𝑥 ٢‬‬ ‫‪i‬‬ ‫‪١‬‬ ‫‪e‬‬

‫‪W‬‬ ‫‪β‬‬ ‫‪١‬‬ ‫‪λx‬‬
‫‪dβ‬‬ ‫𝛽‬ ‫‪١‬‬ ‫‪٢‬‬ ‫‪n‬‬ ‫‪١‬‬ ‫𝛽‬ ‫‪١‬‬
‫‪١‬‬

‫‪𝜆٢ 𝑥 ٢‬‬ ‫‪١‬‬ ‫‪e‬‬

‫‪β‬‬ ‫‪١‬‬ ‫‪λx‬‬
‫‪٢‬‬ ‫𝛽‬ ‫‪١‬‬ ‫‪٢‬‬

‫ﺻﻔﺤﺔ ‪222‬‬ ‫ﻣﺟﻠﺔ ﺍﻟﺩﻧﺎﻧﻳﺭ ) ﺍﻟﻌﺩ‪(٢٤‬‬

 ‫ﺍﺳﺘﺨﺪﺍﻡ ﻁﺮﻳﻘﺔ ﺍﻟﻠﻤﺮﺑﻌﺎﺕ ﺍﻟﺼﻐﺮﻯ ﺍﻟﻤﻮﺯﻭﻧﺔ ﻟﺘﻘﺪﻳﺮ ﺍﻟﺒﻘﺎء ﻟﻠﻨﻮﺫﺝ ﺍﻻﺣﺘﻤﺎﻟﻲ ﺍﻟﻤﺨﺘﻠﻂ )ﺍﻻﺳﻲ –ﻛﺎﻣﺎ ﻣﻦ ﺍﻟﺮﺗﺒﺔ ﺍﻟﺜﺎﻧﻴﺔ(‬

‫‪𝜆٢ 𝑥 ٢‬‬
‫‪١‬‬ ‫‪e‬‬ ‫‪٢ ٢‬‬
‫𝑥 𝜆‬ ‫‪i‬‬ ‫‪١‬‬ ‫‪e‬‬ ‫‪١‬‬ ‫‪e‬‬ ‫‪β‬‬ ‫‪١‬‬ ‫‪λx‬‬
‫‪٢‬‬
‫‪W‬‬ ‫‪β‬‬ ‫‪١‬‬ ‫‪λx‬‬
‫𝛽‬ ‫‪١‬‬ ‫‪٢‬‬ ‫‪n‬‬ ‫‪١‬‬ ‫𝛽‬ ‫‪١‬‬ ‫𝛽‬ ‫‪١‬‬ ‫‪٢‬‬
‫‪١‬‬

‫ﻭﺑﺎﻟﻣﺳﺎﻭﺍﺓ ﻟﻠﺻﻔﺭ ﻧﺣﺻﻝ ﻋﻠﻰ‬

‫‪١‬‬ ‫‪e‬‬ ‫‪𝜆٢ 𝑥 ٢‬‬

‫‪٠‬‬ ‫‪W‬‬ ‫𝛽‬ ‫‪١‬‬ ‫‪λx‬‬
‫𝛽‬ ‫‪١‬‬ ‫‪٢‬‬
‫‪١‬‬

‫‪𝜆٢ 𝑥 ٢‬‬
‫𝛽‬ ‫‪١ ١‬‬ ‫‪e‬‬ ‫‪١‬‬ ‫‪e‬‬ ‫𝛽‬ ‫‪١‬‬ ‫‪λx‬‬
‫‪i‬‬ ‫‪٢‬‬
‫‪n‬‬ ‫‪١‬‬ ‫𝛽‬ ‫‪١‬‬
‫‪٢‬‬

‫‪𝜆٢ 𝑥 ٢‬‬
‫‪١‬‬ ‫‪e‬‬ ‫𝛽‬ ‫‪١‬‬ ‫𝛽‬ ‫‪١‬‬ ‫‪λx‬‬
‫‪١‬‬ ‫‪e‬‬ ‫‪𝜆٢ 𝑥 ٢‬‬ ‫‪i‬‬ ‫‪٢‬‬
‫‪٠‬‬ ‫‪W‬‬ ‫𝛽‬ ‫‪١‬‬ ‫‪λx‬‬
‫𝛽‬ ‫‪١‬‬ ‫‪٢‬‬ ‫‪n‬‬ ‫‪١‬‬ ‫𝛽‬ ‫‪١‬‬
‫‪٢‬‬
‫‪١‬‬

‫‪𝜆٢ 𝑥 ٢‬‬
‫‪١‬‬ ‫‪e‬‬ ‫𝛽‬ ‫‪١‬‬ ‫𝛽‬ ‫‪١‬‬ ‫‪λx‬‬
‫‪٢‬‬
‫‪𝑙𝑒𝑡 𝑢١‬‬ ‫‪٢‬‬
‫𝛽‬ ‫‪١‬‬

‫‪𝜆٢ 𝑥 ٢‬‬
‫‪١‬‬ ‫‪e‬‬ ‫‪١‬‬ ‫‪λx‬‬
‫‪𝛽 ١ e‬‬ ‫‪٢‬‬ ‫‪i‬‬
‫‪٠‬‬ ‫‪W‬‬ ‫‪𝑢١‬‬ ‫‪W‬‬ ‫‪𝑢١‬‬ ‫‪W‬‬ ‫‪𝑢١‬‬ ‫𝟔𝟏 …‬
‫‪𝛽 ١‬‬ ‫𝛽‬ ‫‪١‬‬ ‫‪n‬‬ ‫‪١‬‬
‫‪١‬‬ ‫‪١‬‬ ‫‪١‬‬

‫ﻭﺑﺎﻻﺷﺗﻘﺎﻕ ﻟﻠﻣﻌﻠﻣﺔ ‪ λ‬ﻭﺑﺎﻟﻘﺳﻣﺔ ﻋﻠﻰ ‪ ٢‬ﻧﺣﺻﻝ ﻋﻠﻰ ‪:‬‬

‫‪٢‬‬
‫‪dK‬‬ ‫‪١‬‬ ‫‪e‬‬ ‫‪𝜆٢ 𝑥 ٢‬‬ ‫‪i‬‬
‫‪W‬‬ ‫‪β‬‬ ‫‪١‬‬ ‫‪xλ‬‬
‫‪dλ‬‬ ‫𝛽‬ ‫‪١‬‬ ‫‪٢‬‬ ‫‪n‬‬ ‫‪١‬‬
‫‪١‬‬

‫‪١‬‬ ‫‪e‬‬ ‫‪𝜆٢ 𝑥 ٢‬‬ ‫‪١‬‬

‫‪٠‬‬ ‫‪٠‬‬ ‫‪x‬‬ ‫‪λ𝑥 ٢‬‬ ‫‪β‬‬ ‫‪١‬‬ ‫‪λx‬‬ ‫‪xe‬‬
‫𝛽‬ ‫‪١‬‬ ‫‪٢ β ١‬‬

‫ﻭﺑﺎﻟﻣﺳﺎﻭﺍﺓ ﻟﻠﺻﻔﺭ‬

‫‪٢‬‬
‫‪١‬‬ ‫‪e‬‬ ‫‪𝜆٢ 𝑥 ٢‬‬ ‫‪i‬‬
‫‪٠‬‬ ‫‪W‬‬ ‫‪β‬‬ ‫‪١‬‬ ‫‪λx‬‬
‫𝛽‬ ‫‪١‬‬ ‫‪٢‬‬ ‫‪n‬‬ ‫‪١‬‬
‫‪١‬‬

‫ﺻﻔﺤﺔ ‪223‬‬ ‫ﻣﺟﻠﺔ ﺍﻟﺩﻧﺎﻧﻳﺭ ) ﺍﻟﻌﺩ‪(٢٤‬‬

 ‫ﺍﺳﺘﺨﺪﺍﻡ ﻁﺮﻳﻘﺔ ﺍﻟﻠﻤﺮﺑﻌﺎﺕ ﺍﻟﺼﻐﺮﻯ ﺍﻟﻤﻮﺯﻭﻧﺔ ﻟﺘﻘﺪﻳﺮ ﺍﻟﺒﻘﺎء ﻟﻠﻨﻮﺫﺝ ﺍﻻﺣﺘﻤﺎﻟﻲ ﺍﻟﻤﺨﺘﻠﻂ )ﺍﻻﺳﻲ –ﻛﺎﻣﺎ ﻣﻦ ﺍﻟﺮﺗﺒﺔ ﺍﻟﺜﺎﻧﻴﺔ(‬

‫‪١‬‬ ‫‪e‬‬ ‫‪𝜆٢ 𝑥 ٢‬‬ ‫‪١‬‬

‫‪٠‬‬ ‫‪٠‬‬ ‫‪x‬‬ ‫‪𝜆𝑥 ٢‬‬ ‫‪β‬‬ ‫‪١‬‬ ‫‪λx‬‬ ‫‪xe‬‬
‫𝛽‬ ‫‪١‬‬ ‫‪٢ β ١‬‬

‫‪𝜆٢ 𝑥 ٢‬‬
‫‪xe‬‬ ‫‪β‬‬ ‫‪١‬‬ ‫‪λx‬‬
‫‪١‬‬ ‫‪e‬‬ ‫‪x‬‬ ‫‪𝜆𝑥 ٢‬‬ ‫‪٢‬‬
‫‪𝑙𝑒𝑡 𝑢٢‬‬
‫𝛽‬ ‫‪١‬‬ ‫‪β‬‬ ‫‪١‬‬

‫‪𝜆٢ 𝑥 ٢‬‬
‫‪𝛽 ١‬‬ ‫‪e‬‬ ‫‪١‬‬ ‫‪e‬‬ ‫‪١‬‬ ‫‪λx‬‬ ‫‪i‬‬
‫‪٠‬‬ ‫‪W‬‬ ‫‪𝑢٢‬‬ ‫‪W‬‬ ‫𝑢 ‪٢‬‬ ‫‪W‬‬ ‫𝟕𝟏 … ‪𝑢٢‬‬
‫‪٢‬‬
‫𝛽‬ ‫‪١‬‬ ‫𝛽‬ ‫‪١‬‬ ‫‪n‬‬ ‫‪١‬‬
‫‪١‬‬ ‫‪١‬‬ ‫‪١‬‬

‫ﺍﻟﻣﻌﺎﺩﻻﺕ )‪ (١٧)(١٦‬ﺗﻣﺛﻝ ﺻﻳﻎ ﺧﻁﻳﺔ ﻏﻳﺭ ﻣﻐﻠﻘﺔ ﻭﺍﻟﺗﻲ ﻻ ﻳﻣﻛﻥ ﺣﻠﻬﺎ ﺍﻻ ﺑﺎﺳﺗﺧﺩﺍﻡ ﺍﻟﻁﺭﻳﻘﺔ ﺍﺣﺩﻯ ﺍﻟﻁﺭﺍﺋﻖ ﺍﻟﺗﻛﺭﺍﺭﻳﺔ ﺍﻟﻌﺩﺩﻳﺔ ﻟﺗﻘﺩﻳﺭ‬
‫‪ (β‬ﻭﻣﻘﺩﺭ ﺍﻟﻣﺭﺑﻌﺎﺕ ﺍﻟﺻﻐﺭﻯ ﺍﻟﻣﻭﺯﻭﻧﺔ ﻟﺩﺍﻟﺔ ﺍﻟﺑﻘﺎء ﻛﺎﻷﺗﻲ ‪:‬‬ ‫‪,λ‬‬ ‫)‪,‬‬ ‫ﺍﻟﻣﻌﻠﻣﺎﺕ‬

‫𝑒‬ ‫‪λ٢‬‬ ‫𝑡‬

‫‪s t‬‬ ‫‪١‬‬ ‫‪β‬‬ ‫‪λ‬‬ ‫𝑡‬ ‫𝟖𝟏 …‬
‫‪β‬‬ ‫‪١‬‬ ‫‪٢‬‬

‫‪ .٤‬ﺍﻟﺑﻳﺎﻧﺎﺕ ﺍﻟﺗﻁﺑﻳﻘﻳﺔ‬
‫ﺗﻡ ﺍﻟﺣﺻﻭﻝ ﻋﻠﻰ ﺑﻳﺎﻧﺎﺕ ﻋﻳﻧﺔ ﻣﻥ ﺍﻟﻣﺻﺎﺑﻳﻥ ﺑﺎﻟﻔﺷﻝ ﺍﻟﻛﻠﻭﻱ ﻭﻣﻥ ﻛﻼ ﺍﻟﺟﻧﺳﻳﻥ ﻟﻠﻔﺗﺭﺓ )‪ ٢٠١٥/١/١٧‬ﻟﻐﺎﻳﺔ ‪ (٢٠١٧/١٢/٢٧‬ﻭﺑﻭﺍﻗﻊ‬
‫)‪ (٩١‬ﻣﺻﺎﺏ ﻭﺫﻟﻙ ﺑﻬﺩﻑ ﺗﻁﺑﻳﻘﻬﺎ ﻋﻠﻰ ﺍﻟﺗﻭﺯﻳﻊ ﺍﻟﻣﺧﺗﻠﻁ ) ﺍﻻﺳﻲ _ ﻛﺎﻣﺎ ﻣﻥ ﺍﻟﺩﺭﺟﺔ ﺍﻟﺛﺎﻧﻳﺔ( ﻭﻋﻠﻰ ﻛﻝ ﻣﻥ ﺍﻟﺗﻭﺯﻳﻌﻳﻥ ﺍﻟﻣﻔﺭﺩﻳﻥ‬
‫)ﺍﻻﺳﻲ ( ) ﻛﺎﻣﺎ ﻣﻥ ﺍﻟﺩﺭﺟﺔ ﺍﻟﺛﺎﻧﻳﺔ( ﻭﻣﻥ ﺛﻡ ﺗﻘﺩﻳﺭ ﺩﺍﻟﺔ ﺍﻟﺑﻘﺎء ﺑﺄﺳﺗﺧﺩﺍﻡ ﻁﺭﻳﻘﺔ ﺍﻟﻣﺭﺑﻌﺎﺕ ﺍﻟﺻﻐﺭﻯ ﺍﻟﻣﻭﺯﻭﻧﺔ )𝑠𝑙𝑤( ﺑﺎﻻﻋﺗﻣﺎﺩ ﻋﻠﻰ‬
‫ﺑﺭﻧﺎﻣﺞ ﻛﺗﺏ ﻟﻐﺔ )‪. (Matlab‬‬

‫‪ ١‐٤‬ﺟﻣﻊ ﺍﻟﺑﻳﺎﻧﺎﺕ ﺍﻟﻣﺗﻌﻠﻘﺔ ﺑﺎﻟﺩﺭﺍﺳﺔ‬

‫ﺗﻣﺕ ﻋﻣﻠﻳﺔ ﺟﻣﻊ ﺍﻟﺑﻳﺎﻧﺎﺕ ﺍﻟﺧﺎﺻﺔ ﺑﻭﻓﻳﺎﺕ ﺍﻟﻣﺻﺎﺑﻳﻥ ﺑﺎﻟﻔﺷﻝ ﺍﻟﻛﻠﻭﻱ ﺑﺎﻟﺗﻧﺳﻳﻖ ﻣﻊ ﺷﻌﺑﺔ ﺍﻻﺣﺻﺎء ﺍﻟﺣﻳﺎﺗﻲ ﺿﻣﻥ ﻣﺭﻛﺯ ﺍﻣﺭﺍﺽ‬
‫ﻭﺯﺭﺍﻋﺔ ﺍﻟﻛﻠﻰ ﻓﻘﺩ ﺗﻡ ﺍﻟﺣﺻﻭﻝ ﻋﻠﻰ ﺑﻳﺎﻧﺎﺕ ﻟـ‪٩١‬ﻣﺷﺎﻫﺩﺓ )ﻋﻳﻧﺔ ﻛﺑﻳﺭﺓ ﺍﻟﺣﺟﻡ ( ﺍﻟﻭﻓﻳﺎﺕ ﺍﻟﺣﺎﺻﻠﺔ ﺑﺳﺑﺏ ﻣﺭﺽ ﺍﻟﻔﺷﻝ ﺍﻟﻛﻠﻭﻱ ﻭﻗﺩ‬
‫ﺑﻭﺑﺕ ﻫﺫﻩ ﺍﻟﺑﻳﺎﻧﺎﺕ ﻭﺗﻡ ﺍﻟﺣﺻﻭﻝ ﻋﻠﻰ ﺍﻭﻗﺎﺕ ﺍﻟﺣﻳﺎﺓ ﻟﻠﻣﺭﺿﻰ ﺍﻟﻣﺻﺎﺑﻳﻥ ﺑﺎﻟﻔﺷﻝ ﺍﻟﻛﻠﻭﻱ ﺑﺎﻷﻳﺎﻡ ) 𝑡( ﻭﺫﻟﻙ ﺑﻁﺭﺡ ﺗﺎﺭﻳﺦ ﺍﻻﺻﺎﺑﺔ ﻣﻥ‬
‫ﺗﺎﺭﻳﺦ ﺍﻟﻭﻓﺎﺓ ﻭﻓﻳﻣﺎ ﻳﻠﻲ ﺟﺩﻭﻝ ﻳﺑﻳﻥ ﺍﻭﻗﺎﺕ ﺍﻟﺣﻳﺎﺓ ﻟﻠﻣﺻﺎﺑﻳﻥ ﺑﻣﺭﺽ ﺍﻟﻔﺷﻝ ﺍﻟﻛﻠﻭﻱ‪.‬‬

‫ﺟﺪﻭﻝ )‪ (١‬ﺍﻭﻗﺎﺕ ﺍﻟﺤﻴﺎﺓ ﻟﻠﻤﺮﺿﻰ ﺍﻟﻤﺼﺎﺑﻴﻦ ﺑﺎﻟﻔﺸﻞ ﺍﻟﻜﻠﻮﻱ ‪.‬‬

‫‪I‬‬ ‫‪t‬‬ ‫‪S_real‬‬ ‫‪i‬‬ ‫‪t‬‬ ‫‪S_real‬‬ ‫‪i‬‬ ‫‪t‬‬ ‫‪S_real‬‬
‫‪١‬‬ ‫‪١‬‬ ‫‪٠.٩٧‬‬ ‫‪٣١‬‬ ‫‪٢‬‬ ‫‪٠.٩٥‬‬ ‫‪٦١‬‬ ‫‪٧‬‬ ‫‪٠.٨٣‬‬
‫‪٢‬‬ ‫‪١‬‬ ‫‪٠.٩٧‬‬ ‫‪٣٢‬‬ ‫‪٢‬‬ ‫‪٠.٩٥‬‬ ‫‪٦٢‬‬ ‫‪٧‬‬ ‫‪٠.٨٣‬‬
‫‪٣‬‬ ‫‪١‬‬ ‫‪٠.٩٧‬‬ ‫‪٣٣‬‬ ‫‪٣‬‬ ‫‪٠.٩٢‬‬ ‫‪٦٣‬‬ ‫‪٧‬‬ ‫‪٠.٨٣‬‬
‫‪٤‬‬ ‫‪١‬‬ ‫‪٠.٩٧‬‬ ‫‪٣٤‬‬ ‫‪٣‬‬ ‫‪٠.٩٢‬‬ ‫‪٦٤‬‬ ‫‪٧‬‬ ‫‪٠.٨٣‬‬

‫ﺻﻔﺤﺔ ‪224‬‬ ‫ﻣﺟﻠﺔ ﺍﻟﺩﻧﺎﻧﻳﺭ ) ﺍﻟﻌﺩ‪(٢٤‬‬

 ‫ﺍﺳﺘﺨﺪﺍﻡ ﻁﺮﻳﻘﺔ ﺍﻟﻠﻤﺮﺑﻌﺎﺕ ﺍﻟﺼﻐﺮﻯ ﺍﻟﻤﻮﺯﻭﻧﺔ ﻟﺘﻘﺪﻳﺮ ﺍﻟﺒﻘﺎء ﻟﻠﻨﻮﺫﺝ ﺍﻻﺣﺘﻤﺎﻟﻲ ﺍﻟﻤﺨﺘﻠﻂ )ﺍﻻﺳﻲ –ﻛﺎﻣﺎ ﻣﻦ ﺍﻟﺮﺗﺒﺔ ﺍﻟﺜﺎﻧﻴﺔ(‬

‫‪٥‬‬ ‫‪١‬‬ ‫‪٠.٩٧‬‬ ‫‪٣٥‬‬ ‫‪٣‬‬ ‫‪٠.٩٢‬‬ ‫‪٦٥‬‬ ‫‪٧‬‬ ‫‪٠.٨٣‬‬

‫‪٦‬‬ ‫‪١‬‬ ‫‪٠.٩٧‬‬ ‫‪٣٦‬‬ ‫‪٣‬‬ ‫‪٠.٩٢‬‬ ‫‪٦٦‬‬ ‫‪٧‬‬ ‫‪٠.٨٣‬‬
‫‪٧‬‬ ‫‪١‬‬ ‫‪٠.٩٧‬‬ ‫‪٣٧‬‬ ‫‪٣‬‬ ‫‪٠.٩٢‬‬ ‫‪٦٧‬‬ ‫‪٨‬‬ ‫‪٠.٨١‬‬
‫‪٨‬‬ ‫‪١‬‬ ‫‪٠.٩٧‬‬ ‫‪٣٨‬‬ ‫‪٣‬‬ ‫‪٠.٩٢‬‬ ‫‪٦٨‬‬ ‫‪٨‬‬ ‫‪٠.٨١‬‬
‫‪٩‬‬ ‫‪١‬‬ ‫‪٠.٩٧‬‬ ‫‪٣٩‬‬ ‫‪٣‬‬ ‫‪٠.٩٢‬‬ ‫‪٦٩‬‬ ‫‪٩‬‬ ‫‪٠.٧٩‬‬
‫‪١٠‬‬ ‫‪١‬‬ ‫‪٠.٩٧‬‬ ‫‪٤٠‬‬ ‫‪٣‬‬ ‫‪٠.٩٢‬‬ ‫‪٧٠‬‬ ‫‪٩‬‬ ‫‪٠.٧٩‬‬
‫‪١١‬‬ ‫‪١‬‬ ‫‪٠.٩٧‬‬ ‫‪٤١‬‬ ‫‪٣‬‬ ‫‪٠.٩٢‬‬ ‫‪٧١‬‬ ‫‪١٠‬‬ ‫‪٠.٧٨‬‬
‫‪١٢‬‬ ‫‪١‬‬ ‫‪٠.٩٧‬‬ ‫‪٤٢‬‬ ‫‪٣‬‬ ‫‪٠.٩٢‬‬ ‫‪٧٢‬‬ ‫‪١١‬‬ ‫‪٠.٧٦‬‬
‫‪١٣‬‬ ‫‪١‬‬ ‫‪٠.٩٧‬‬ ‫‪٤٣‬‬ ‫‪٣‬‬ ‫‪٠.٩٢‬‬ ‫‪٧٣‬‬ ‫‪١١‬‬ ‫‪٠.٧٦‬‬
‫‪١٤‬‬ ‫‪١‬‬ ‫‪٠.٩٧‬‬ ‫‪٤٤‬‬ ‫‪٣‬‬ ‫‪٠.٩٢‬‬ ‫‪٧٤‬‬ ‫‪١٢‬‬ ‫‪٠.٧٤‬‬
‫‪١٥‬‬ ‫‪١‬‬ ‫‪٠.٩٧‬‬ ‫‪٤٥‬‬ ‫‪٤‬‬ ‫‪٠.٩٠‬‬ ‫‪٧٥‬‬ ‫‪١٢‬‬ ‫‪٠.٧٤‬‬
‫‪١٦‬‬ ‫‪١‬‬ ‫‪٠.٩٧‬‬ ‫‪٤٦‬‬ ‫‪٤‬‬ ‫‪٠.٩٠‬‬ ‫‪٧٦‬‬ ‫‪١٢‬‬ ‫‪٠.٧٤‬‬
‫‪١٧‬‬ ‫‪١‬‬ ‫‪٠.٩٧‬‬ ‫‪٤٧‬‬ ‫‪٤‬‬ ‫‪٠.٩٠‬‬ ‫‪٧٧‬‬ ‫‪١٣‬‬ ‫‪٠.٧٢‬‬
‫‪١٨‬‬ ‫‪١‬‬ ‫‪٠.٩٧‬‬ ‫‪٤٨‬‬ ‫‪٤‬‬ ‫‪٠.٩٠‬‬ ‫‪٧٨‬‬ ‫‪١٤‬‬ ‫‪٠.٧١‬‬
‫‪١٩‬‬ ‫‪١‬‬ ‫‪٠.٩٧‬‬ ‫‪٤٩‬‬ ‫‪٥‬‬ ‫‪٠.٨٨‬‬ ‫‪٧٩‬‬ ‫‪١٤‬‬ ‫‪٠.٧١‬‬
‫‪٢٠‬‬ ‫‪٢‬‬ ‫‪٠.٩٥‬‬ ‫‪٥٠‬‬ ‫‪٥‬‬ ‫‪٠.٨٨‬‬ ‫‪٨٠‬‬ ‫‪١٧‬‬ ‫‪٠.٦٦‬‬
‫‪٢١‬‬ ‫‪٢‬‬ ‫‪٠.٩٥‬‬ ‫‪٥١‬‬ ‫‪٥‬‬ ‫‪٠.٨٨‬‬ ‫‪٨١‬‬ ‫‪١٨‬‬ ‫‪٠.٦٤‬‬
‫‪٢٢‬‬ ‫‪٢‬‬ ‫‪٠.٩٥‬‬ ‫‪٥٢‬‬ ‫‪٥‬‬ ‫‪٠.٨٨‬‬ ‫‪٨٢‬‬ ‫‪١٨‬‬ ‫‪٠.٦٤‬‬
‫‪٢٣‬‬ ‫‪٢‬‬ ‫‪٠.٩٥‬‬ ‫‪٥٣‬‬ ‫‪٥‬‬ ‫‪٠.٨٨‬‬ ‫‪٨٣‬‬ ‫‪١٨‬‬ ‫‪٠.٦٤‬‬
‫‪٢٤‬‬ ‫‪٢‬‬ ‫‪٠.٩٥‬‬ ‫‪٥٤‬‬ ‫‪٥‬‬ ‫‪٠.٨٨‬‬ ‫‪٨٤‬‬ ‫‪١٩‬‬ ‫‪٠.٦٣‬‬
‫‪٢٥‬‬ ‫‪٢‬‬ ‫‪٠.٩٥‬‬ ‫‪٥٥‬‬ ‫‪٥‬‬ ‫‪٠.٨٨‬‬ ‫‪٨٥‬‬ ‫‪١٩‬‬ ‫‪٠.٦٣‬‬
‫‪٢٦‬‬ ‫‪٢‬‬ ‫‪٠.٩٥‬‬ ‫‪٥٦‬‬ ‫‪٥‬‬ ‫‪٠.٨٨‬‬ ‫‪٨٦‬‬ ‫‪٢٠‬‬ ‫‪٠.٦١‬‬
‫‪٢٧‬‬ ‫‪٢‬‬ ‫‪٠.٩٥‬‬ ‫‪٥٧‬‬ ‫‪٥‬‬ ‫‪٠.٨٨‬‬ ‫‪٨٧‬‬ ‫‪٢٠‬‬ ‫‪٠.٦١‬‬
‫‪٢٨‬‬ ‫‪٢‬‬ ‫‪٠.٩٥‬‬ ‫‪٥٨‬‬ ‫‪٦‬‬ ‫‪٠.٨٥‬‬ ‫‪٨٨‬‬ ‫‪٢٥‬‬ ‫‪٠.٥٥‬‬
‫‪٢٩‬‬ ‫‪٢‬‬ ‫‪٠.٩٥‬‬ ‫‪٥٩‬‬ ‫‪٦‬‬ ‫‪٠.٨٥‬‬ ‫‪٨٩‬‬ ‫‪٢٥‬‬ ‫‪٠.٥٥‬‬
‫‪٣٠‬‬ ‫‪٢‬‬ ‫‪٠.٩٥‬‬ ‫‪٦٠‬‬ ‫‪٧‬‬ ‫‪٠.٨٣‬‬ ‫‪٩٠‬‬ ‫‪٢٦‬‬ ‫‪٠.٥٤‬‬
‫‪٩١‬‬ ‫‪٢٨‬‬ ‫‪٠.٥١‬‬
‫ﺟﺩﻭﻝ )‪ (٢‬ﻧﺗﺎﺋﺞ ﺍﻟﻣﺅﺷﺭﺍﺕ ﺍﻻﺣﺻﺎﺋﻳﺔ ﻟﻠﻌﻳﻧﺔ‪:‬‬

‫‪Statistic‬‬ ‫‪Value‬‬

‫‪Sample Size‬‬ ‫‪٩١‬‬

‫‪Mean‬‬ ‫‪٦.٥٨٢٤١٨‬‬

‫‪Variance‬‬ ‫‪٤٥.٠٩٠٣٥‬‬

‫‪Std. Deviation‬‬ ‫‪٦.٧١٤٩٣٥‬‬

‫‪Coef. of Variation‬‬ ‫‪٠.٩١٨٤٧‬‬ ‫‪٢.٤‬‬

‫ﺻﻔﺤﺔ ‪225‬‬ ‫ﻣﺟﻠﺔ ﺍﻟﺩﻧﺎﻧﻳﺭ ) ﺍﻟﻌﺩ‪(٢٤‬‬

 ‫ﺍﺳﺘﺨﺪﺍﻡ ﻁﺮﻳﻘﺔ ﺍﻟﻠﻤﺮﺑﻌﺎﺕ ﺍﻟﺼﻐﺮﻯ ﺍﻟﻤﻮﺯﻭﻧﺔ ﻟﺘﻘﺪﻳﺮ ﺍﻟﺒﻘﺎء ﻟﻠﻨﻮﺫﺝ ﺍﻻﺣﺘﻤﺎﻟﻲ ﺍﻟﻤﺨﺘﻠﻂ )ﺍﻻﺳﻲ –ﻛﺎﻣﺎ ﻣﻦ ﺍﻟﺮﺗﺒﺔ ﺍﻟﺜﺎﻧﻴﺔ(‬

‫)‪(Goodness of Fit‬‬ ‫ﺇﺧﺗﺑﺎﺭ ﺣﺳﻥ ﺍﻟﻣﻁﺎﺑﻘﺔ‬

‫ﻠ )ﺍﻻﺳﻲ _ ﻛﺎﻣﺎ ﻣﻥ ﺍﻟﺭﺗﺑﺔ ﺍﻟﺛﺎﻧﻳﺔ( ﺗ اﺳ ﻌ ﺎل ﻣ ﺎراﻛ ﻲ)‪(AIC‬‬ ‫ن اﻟ ﺎﻧﺎت ﻓﻲ اﻟ ول )‪(٣‐١٢‬ﺗ زع اﻟ زﻊ اﻟ‬ ‫ﻟ ﻌ ﻓﺔ‬

‫ﺢ )‪ (AICC‬وﻣ ﺎرﺑﻳﺯﺍﻛﻳﻛﻲ )‪(BIC‬ﻻﺧﺗﺑﺎﺭ ﺣﺳﻥ ﺍﻟﻣﻁﺎﺑﻘﺔ ﻟﻠﺑﻳﺎﻧﺎﺕ ﻭﺗﺑﻳﻥ ﺍﻧﻬﺎ ﺗﺗﻭﺯﻉ ﻭﻓﻘﺎ ﻟﻠﻧﻣﻭﺫﺝ ﺍﻻﺣﺗﻣﺎﻟﻲ‬ ‫ﻭﻣ ﺎراﻛ ﻲ اﻟ‬

‫ﺍﻟﻣﺧﺗﻠﻁ )ﺍﻻﺳﻲ _ ﻛﺎﻣﺎ ﻣﻥ ﺍﻟﺭﺗﺑﺔ ﺍﻟﺛﺎﻧﻳﺔ( ﻭﻗﺩ ﺗﻡ ﻗﺑﻭﻝ ﺍﻟﻔﺭﺿﻳﺔ ﺍﻟﻘﺎﺋﻠﺔ ﺑﺄﻥ ﺍﻟﺑﻳﺎﻧﺎﺕ ﺗﺗﻭﺯﻉ ﺍﻟﻧﻣﻭﺫﺝ ﺍﻻﺣﺗﻣﺎﻟﻲ ﺍﻟﻣﺧﺗﻠﻁ )ﺍﻻﺳﻲ _‬
‫ﻛﺎﻣﺎ ﻣﻥ ﺍﻟﺭﺗﺑﺔ ﺍﻟﺛﺎﻧﻳﺔ( ‪:‬‬
‫‪H٠ : the data are EG.‬‬
‫‪H١ : the data are not EG‬‬

‫ﻭﻳﻭﺿﺢ ﺍﻟﺟﺩﻭﻝ ﺍﺩﻧﺎﻩ ﻧﺗﺎﺋﺞ ﺍﺧﺗﺑﺎﺭ ﺣﺳﻥ ﺍﻟﻣﻁﺎﺑﻘﺔ ﻣﻥ ﺧﻼﻝ ﻭﺍﻟﺗﻲ ﺗﻡ ﺍﻟﺣﺻﻭﻝ ﻋﻠﻳﻬﺎ ﻣﻥ ﺧﻼﻝ ﺑﺭﻧﺎﻣﺞ ﺗﻡ ﻛﺗﺎﺑﺗﻪ ﺑﺎﺳﺗﺧﺩﺍﻡ ﻟﻐﺔ ﻣﺎﺗﻼﺏ‬

‫𝜒‪Chi-Squared‬‬ ‫∑‬ ‫𝜒~‬ ‫)‪ (٢٠١٢‬ﻭﺑﺄﺳﺗﺧﺩﺍﻡ ﻗﺎﻧﻭﻥ‬

‫ﺟﺩﻭﻝ )‪ (٣‬ﻧﺗﺎﺋﺞ ﺍﺧﺗﺑﺎﺭ ﺣﺳﻥ ﺍﻟﻣﻁﺎﺑﻘﺔ‬

‫‪Distribution‬‬ ‫‪p‐value‬‬ ‫‪𝜒٢‬‬ ‫‪𝜒٢‬‬ ‫‪decision‬‬

‫‪Exponential‬‬ ‫‪accept H٠‬‬

‫‪٠.٣٦‬‬ ‫‪٤.٣٧‬‬ ‫‪٩.٤٩‬‬

‫‪Gamma‬‬ ‫‪accept H٠‬‬

‫‪٠.٠٩‬‬ ‫‪٦.٥٣‬‬ ‫‪٧.٨٢‬‬

‫‪EG‬‬ ‫‪accept H٠‬‬

‫‪٠.٠٨‬‬ ‫‪٦.٦٤‬‬ ‫‪٧.٨٢‬‬

‫ﻣ اﻟ وﻟ ﺔ‬ ‫ﺔ اﻛ‬ ‫ﻣﻌ ﺔ )‪ (٠.٠٥‬و ﺔ 𝟐𝝌 اﻟ‬ ‫ﻣ ﻣ‬ ‫‪ ‬ﻧﻼﺣ ﻣ اﻟ ول اﻋﻼﻩ ان اﻛ‬

‫اذاً ﺗﻘ ﻞ ﻓ ﺿ ﺔ اﻟﻌ م اﻟﻘﺎﺋﻠﺔ ﺄن اﻟ ﺎﻧﺎت ﺗ زع ﺗ زﻊ )اﻻﺳﻲ –ﻛﺎﻣﺎ ﻣ اﻟ رﺟﺔ اﻟ ﺎﻧ ﺔ(‪.‬‬

‫]‪[٣‬‬
‫‪ ٣.٤‬ﻣﻌﺎﻳﻳﺭ ﺍﺧﺗﻳﺎﺭ ﺍﻓﺿﻝ ﺗﻭﺯﻳﻊ‪:‬‬
‫‪Criteria for choosing the best distribution‬‬

‫‪(AIC)Akaike's Test‬‬ ‫‪ -١‬اﺧ ﺎر اﻛ ﻲ‪:‬‬

‫ﺎ ﺎﺗﻲ‪:‬‬ ‫وﺗ ن ﺻ ﻐﺔ اﺣ ﺎءة ﻣ ﺎر اﻛ ﻲ)‪(AIC‬‬

‫‪AIC‬‬ ‫‪٢L θ ∖ x‬‬ ‫‪٢p‬‬ ‫)‪…(١٩‬‬

‫ﺻﻔﺤﺔ ‪226‬‬ ‫ﻣﺟﻠﺔ ﺍﻟﺩﻧﺎﻧﻳﺭ ) ﺍﻟﻌﺩ‪(٢٤‬‬

 ‫ﺍﺳﺘﺨﺪﺍﻡ ﻁﺮﻳﻘﺔ ﺍﻟﻠﻤﺮﺑﻌﺎﺕ ﺍﻟﺼﻐﺮﻯ ﺍﻟﻤﻮﺯﻭﻧﺔ ﻟﺘﻘﺪﻳﺮ ﺍﻟﺒﻘﺎء ﻟﻠﻨﻮﺫﺝ ﺍﻻﺣﺘﻤﺎﻟﻲ ﺍﻟﻤﺨﺘﻠﻂ )ﺍﻻﺳﻲ –ﻛﺎﻣﺎ ﻣﻦ ﺍﻟﺮﺗﺒﺔ ﺍﻟﺜﺎﻧﻴﺔ(‬

‫𝐱 ∖ 𝛉 𝐋 ‪ :‬ﺗ ﻞ ﻟ ﻏﺎر داﻟﺔ اﻟ ﺟ ﺢ ‪ Log Likelihood Function‬ﻟ ﺎﻧﺎت اﻟﻌ ﺔ‪.‬‬

‫)‪Akaike Information Criterion(AICc‬‬ ‫ﺢ‬ ‫‪ -٢‬اﺧ ﺎر اﻛ ﻲ اﻟ‬

‫ﻗﺩﻡ ﻛﻝ ﻣﻥ )‪(١٩٩٣) (Brockwell and Davis‬ﻣﻘﺘﺮﺡ ﻟﺘﺼﺤﻴﺢ ﺣﺎﻟﺔ ﺍﻟﺘﺤﻴﺰ ﻓﻲ اﺧ ﺎر اﻛ ﻲ)‪(AIC‬ﻣﻥ ﺧﻼﻝ ﺃﺿﺎﻓﺔ ﺍﻟﻣﻘﺩﺍﺭ‬
‫‪٢‬‬ ‫‪١‬‬ ‫‪٢‬‬
‫ﺍﻟﻰ ﻗﻳﻣﺔ اﻛ ﻲ)‪ (AIC‬ﻟﻳﻛﻭﻥ‬
‫‪٢‬‬

‫)‪… (٢٠‬‬
‫‪٢‬‬ ‫‪١‬‬ ‫‪٢‬‬
‫‪AICc‬‬ ‫‪nlog σ٢‬‬ ‫‪٢k‬‬
‫‪٢‬‬

‫‪(BIC)Bayes Akaike's Test‬‬ ‫‪ -٣‬اﺧ ﺎر ﺑ اﻛ ﻲ‬

‫اﻟ ﺎ ﻘﺔ )‪ (GOF‬وﺗ ن ﺻ ﻐﺔ ﻫ ا اﻻﺧ ﺎرﻫﻲ ‪:‬‬ ‫م ﺎﺧ ﺎر ﻟ‬ ‫ﻣ ﺎر‬

‫‪BIC‬‬ ‫‪٢L θ ∖ x‬‬ ‫‪plog n‬‬ ‫)‪… (٢١‬‬

‫و ان ‪:‬‬

‫‪ L θ ∖ x‬ﺗﻤﺜﻞ ﻟﻮﻏﺎﺭﻳﺘﻢ ﺩﺍﻟﺔ ﺍﻟﺘﺮﺟﻴﺢ )‪ (LogLikelihoodFunction‬ﻟﻤﺸﺎﻫﺪﺍﺕ ﺑﻴﺎﻧﺎﺕ ﺍﻟﻌﻴﻨﺔ‪.‬‬

‫ﺔ‪.‬‬ ‫‪ :p‬ﺗ ﻞ ﻋ د اﻟ ﻌﻠ ﺎت ﻓﻲ داﻟﺔ اﻟ زﻊ اﻻﺣ ﺎﻟ ﺔ اﻟ‬

‫اﻟﻌ ﺔ‬ ‫‪n‬ﺗ ﻞﺣ‬

‫ﻭﺍﻟﺟﺩﻭﻝ ﺍﺩﻧﺎﻩ ﻳﻅﻬﺭ ﻧﺗﺎﺋﺞ ﻣﻌﺎﻳﻳﺭ ﺍﻟﻣﻔﺎﺿﻠﺔ ﺍﻟﺗﻲ ﺗﻡ ﺍﺳﺗﺧﺩﺍﻣﻬﺎ ﻭﻳﺗﺑﻳﻥ ﻣﻧﻬﺎ ﺍﻓﺿﻠﻳﺔ ﺍﻟﻧﻣﻭﺫﺝ ﺍﻟﻣﺧﺗﻠﻁ ) ﺍﻻﺳﻲ – ﻛﺎﻣﺎ ﻣﻥ ﺍﻟﺭﺗﺑﺔ ﺍﻟﺛﺎﻧﻳﺔ (‬
‫ﻋﻠﻰ ﻛﻝ ﻣﻥ ﺍﻟﺗﻭﺯﻳﻌﻳﻥ ﺍﻟﻣﻔﺭﺩﻳﻥ )ﺍﻻﺳﻲ(‪ )،‬ﻛﺎﻣﺎ ﻣﻥ ﺍﻟﺭﺗﺑﺔ ﺍﻟﺛﺎﻧﻳﺔ(‬

‫ﺟﺩﻭﻝ )‪ (٤‬ﻣﻌﺎﻳﻳﺭ ﺍﻟﻣﻔﺎﺿﻠﺔ ﺑﻳﻥ ﺍﻟﺗﻭﺯﻳﻌﺎﺕ‬

‫‪Parameter estimation‬‬
‫‪distribution‬‬ ‫‪AIC‬‬ ‫‪AICc‬‬ ‫‪BIC‬‬
‫𝛽‬ ‫𝜆‬
‫‪exponential‬‬ ‫‪٦.٥٨٢‬‬ ‫‪٥٨٠.٥٢٣١‬‬ ‫‪٥٨٠.٥٦٨‬‬ ‫‪٥٨٣.٠٣٣٩‬‬

‫‪Gamma‬‬ ‫‪٥.٦٩٢‬‬ ‫‪٦٣٧.٧٥٦‬‬ ‫‪٦٣٧.٨٠٠٩‬‬ ‫‪٦٤٠.٢٦٦٩‬‬

‫‪EG‬‬ ‫‪١.٥١‬‬ ‫‪٠.٠٤٦‬‬ ‫‪٥٣٤.٠٥٩٦‬‬ ‫‪٥٣٤.١٩٥٩‬‬ ‫‪٥٣٩.٠٨١٣‬‬

‫‪ ٤.٤‬ﻧﺗﺎﺋﺞ ﺑﺭﻧﺎﻣﺞ ﺍﻟﺑﻳﺎﻧﺎﺕ ﺍﻟﺣﻘﻳﻘﻳﺔ‬

‫ﺻﻔﺤﺔ ‪227‬‬ ‫ﻣﺟﻠﺔ ﺍﻟﺩﻧﺎﻧﻳﺭ ) ﺍﻟﻌﺩ‪(٢٤‬‬

 ‫ﺍﺳﺘﺨﺪﺍﻡ ﻁﺮﻳﻘﺔ ﺍﻟﻠﻤﺮﺑﻌﺎﺕ ﺍﻟﺼﻐﺮﻯ ﺍﻟﻤﻮﺯﻭﻧﺔ ﻟﺘﻘﺪﻳﺮ ﺍﻟﺒﻘﺎء ﻟﻠﻨﻮﺫﺝ ﺍﻻﺣﺘﻤﺎﻟﻲ ﺍﻟﻤﺨﺘﻠﻂ )ﺍﻻﺳﻲ –ﻛﺎﻣﺎ ﻣﻦ ﺍﻟﺮﺗﺒﺔ ﺍﻟﺜﺎﻧﻴﺔ(‬

‫ﺟﺩﻭﻝ )‪ (٥‬ﻣﻘﺩﺭ ﺍﻟﻣﺭﺑﻌﺎﺕ ﺍﻟﺻﻐﺭﻯ ﺍﻟﻣﻭﺯﻭﻧﺔ ﻟﺩﺍﻟﺔ ﺍﻟﺑﻘﺎء ﻟﻠﺑﻳﺎﻧﺎﺕ ﺍﻟﺣﻘﻳﻘﻳــــــﺔ‪.‬‬

‫‪t‬‬ ‫‪S_real‬‬ ‫‪F‬‬ ‫‪t‬‬ ‫‪S_real‬‬ ‫‪F‬‬ ‫‪t‬‬ ‫‪S_real‬‬ ‫‪F‬‬ ‫‪T‬‬ ‫‪S_real‬‬ ‫‪F‬‬

‫‪١‬‬ ‫‪٠.٩٧‬‬ ‫‪٠.٠٣‬‬ ‫‪٢‬‬ ‫‪٠.٩٥‬‬ ‫‪٠.٠٥‬‬ ‫‪٤‬‬ ‫‪٠.٩٠‬‬ ‫‪٠.١٠‬‬ ‫‪٩‬‬ ‫‪٠.٧٩‬‬ ‫‪٠.٢١‬‬

‫‪١‬‬ ‫‪٠.٩٧‬‬ ‫‪٠.٠٣‬‬ ‫‪٢‬‬ ‫‪٠.٩٥‬‬ ‫‪٠.٠٥‬‬ ‫‪٤‬‬ ‫‪٠.٩٠‬‬ ‫‪٠.١٠‬‬ ‫‪١٠‬‬ ‫‪٠.٧٨‬‬ ‫‪٠.٢٢‬‬

‫‪١‬‬ ‫‪٠.٩٧‬‬ ‫‪٠.٠٣‬‬ ‫‪٢‬‬ ‫‪٠.٩٥‬‬ ‫‪٠.٠٥‬‬ ‫‪٥‬‬ ‫‪٠.٨٨‬‬ ‫‪٠.١٢‬‬ ‫‪١١‬‬ ‫‪٠.٧٦‬‬ ‫‪٠.٢٤‬‬

‫‪١‬‬ ‫‪٠.٩٧‬‬ ‫‪٠.٠٣‬‬ ‫‪٢‬‬ ‫‪٠.٩٥‬‬ ‫‪٠.٠٥‬‬ ‫‪٥‬‬ ‫‪٠.٨٨‬‬ ‫‪٠.١٢‬‬ ‫‪١١‬‬ ‫‪٠.٧٦‬‬ ‫‪٠.٢٤‬‬

‫‪١‬‬ ‫‪٠.٩٧‬‬ ‫‪٠.٠٣‬‬ ‫‪٢‬‬ ‫‪٠.٩٥‬‬ ‫‪٠.٠٥‬‬ ‫‪٥‬‬ ‫‪٠.٨٨‬‬ ‫‪٠.١٢‬‬ ‫‪١٢‬‬ ‫‪٠.٧٤‬‬ ‫‪٠.٢٦‬‬

‫‪١‬‬ ‫‪٠.٩٧‬‬ ‫‪٠.٠٣‬‬ ‫‪٢‬‬ ‫‪٠.٩٥‬‬ ‫‪٠.٠٥‬‬ ‫‪٥‬‬ ‫‪٠.٨٨‬‬ ‫‪٠.١٢‬‬ ‫‪١٢‬‬ ‫‪٠.٧٤‬‬ ‫‪٠.٢٦‬‬

‫‪١‬‬ ‫‪٠.٩٧‬‬ ‫‪٠.٠٣‬‬ ‫‪٢‬‬ ‫‪٠.٩٥‬‬ ‫‪٠.٠٥‬‬ ‫‪٥‬‬ ‫‪٠.٨٨‬‬ ‫‪٠.١٢‬‬ ‫‪١٢‬‬ ‫‪٠.٧٤‬‬ ‫‪٠.٢٦‬‬

‫‪١‬‬ ‫‪٠.٩٧‬‬ ‫‪٠.٠٣‬‬ ‫‪٢‬‬ ‫‪٠.٩٥‬‬ ‫‪٠.٠٥‬‬ ‫‪٥‬‬ ‫‪٠.٨٨‬‬ ‫‪٠.١٢‬‬ ‫‪١٣‬‬ ‫‪٠.٧٢‬‬ ‫‪٠.٢٨‬‬

‫‪١‬‬ ‫‪٠.٩٧‬‬ ‫‪٠.٠٣‬‬ ‫‪٢‬‬ ‫‪٠.٩٥‬‬ ‫‪٠.٠٥‬‬ ‫‪٥‬‬ ‫‪٠.٨٨‬‬ ‫‪٠.١٢‬‬ ‫‪١٤‬‬ ‫‪٠.٧١‬‬ ‫‪٠.٢٩‬‬

‫‪١‬‬ ‫‪٠.٩٧‬‬ ‫‪٠.٠٣‬‬ ‫‪٣‬‬ ‫‪٠.٩٢‬‬ ‫‪٠.٠٨‬‬ ‫‪٥‬‬ ‫‪٠.٨٨‬‬ ‫‪٠.١٢‬‬ ‫‪١٤‬‬ ‫‪٠.٧١‬‬ ‫‪٠.٢٩‬‬

‫‪١‬‬ ‫‪٠.٩٧‬‬ ‫‪٠.٠٣‬‬ ‫‪٣‬‬ ‫‪٠.٩٢‬‬ ‫‪٠.٠٨‬‬ ‫‪٥‬‬ ‫‪٠.٨٨‬‬ ‫‪٠.١٢‬‬ ‫‪١٧‬‬ ‫‪٠.٦٦‬‬ ‫‪٠.٣٤‬‬

‫‪١‬‬ ‫‪٠.٩٧‬‬ ‫‪٠.٠٣‬‬ ‫‪٣‬‬ ‫‪٠.٩٢‬‬ ‫‪٠.٠٨‬‬ ‫‪٦‬‬ ‫‪٠.٨٥‬‬ ‫‪٠.١٥‬‬ ‫‪١٨‬‬ ‫‪٠.٦٤‬‬ ‫‪٠.٣٦‬‬

‫‪١‬‬ ‫‪٠.٩٧‬‬ ‫‪٠.٠٣‬‬ ‫‪٣‬‬ ‫‪٠.٩٢‬‬ ‫‪٠.٠٨‬‬ ‫‪٦‬‬ ‫‪٠.٨٥‬‬ ‫‪٠.١٥‬‬ ‫‪١٨‬‬ ‫‪٠.٦٤‬‬ ‫‪٠.٣٦‬‬

‫‪١‬‬ ‫‪٠.٩٧‬‬ ‫‪٠.٠٣‬‬ ‫‪٣‬‬ ‫‪٠.٩٢‬‬ ‫‪٠.٠٨‬‬ ‫‪٧‬‬ ‫‪٠.٨٣‬‬ ‫‪٠.١٧‬‬ ‫‪١٨‬‬ ‫‪٠.٦٤‬‬ ‫‪٠.٣٦‬‬

‫‪١‬‬ ‫‪٠.٩٧‬‬ ‫‪٠.٠٣‬‬ ‫‪٣‬‬ ‫‪٠.٩٢‬‬ ‫‪٠.٠٨‬‬ ‫‪٧‬‬ ‫‪٠.٨٣‬‬ ‫‪٠.١٧‬‬ ‫‪١٩‬‬ ‫‪٠.٦٣‬‬ ‫‪٠.٣٧‬‬

‫‪١‬‬ ‫‪٠.٩٧‬‬ ‫‪٠.٠٣‬‬ ‫‪٣‬‬ ‫‪٠.٩٢‬‬ ‫‪٠.٠٨‬‬ ‫‪٧‬‬ ‫‪٠.٨٣‬‬ ‫‪٠.١٧‬‬ ‫‪١٩‬‬ ‫‪٠.٦٣‬‬ ‫‪٠.٣٧‬‬

‫‪١‬‬ ‫‪٠.٩٧‬‬ ‫‪٠.٠٣‬‬ ‫‪٣‬‬ ‫‪٠.٩٢‬‬ ‫‪٠.٠٨‬‬ ‫‪٧‬‬ ‫‪٠.٨٣‬‬ ‫‪٠.١٧‬‬ ‫‪٢٠‬‬ ‫‪٠.٦١‬‬ ‫‪٠.٣٩‬‬

‫‪١‬‬ ‫‪٠.٩٧‬‬ ‫‪٠.٠٣‬‬ ‫‪٣‬‬ ‫‪٠.٩٢‬‬ ‫‪٠.٠٨‬‬ ‫‪٧‬‬ ‫‪٠.٨٣‬‬ ‫‪٠.١٧‬‬ ‫‪٢٠‬‬ ‫‪٠.٦١‬‬ ‫‪٠.٣٩‬‬

‫‪١‬‬ ‫‪٠.٩٧‬‬ ‫‪٠.٠٣‬‬ ‫‪٣‬‬ ‫‪٠.٩٢‬‬ ‫‪٠.٠٨‬‬ ‫‪٧‬‬ ‫‪٠.٨٣‬‬ ‫‪٠.١٧‬‬ ‫‪٢٥‬‬ ‫‪٠.٥٥‬‬ ‫‪٠.٤٥‬‬

‫‪٢‬‬ ‫‪٠.٩٥‬‬ ‫‪٠.٠٥‬‬ ‫‪٣‬‬ ‫‪٠.٩٢‬‬ ‫‪٠.٠٨‬‬ ‫‪٧‬‬ ‫‪٠.٨٣‬‬ ‫‪٠.١٧‬‬ ‫‪٢٥‬‬ ‫‪٠.٥٥‬‬ ‫‪٠.٤٥‬‬

‫‪٢‬‬ ‫‪٠.٩٥‬‬ ‫‪٠.٠٥‬‬ ‫‪٣‬‬ ‫‪٠.٩٢‬‬ ‫‪٠.٠٨‬‬ ‫‪٨‬‬ ‫‪٠.٨١‬‬ ‫‪٠.١٩‬‬ ‫‪٢٦‬‬ ‫‪٠.٥٤‬‬ ‫‪٠.٤٦‬‬

‫‪٢‬‬ ‫‪٠.٩٥‬‬ ‫‪٠.٠٥‬‬ ‫‪٤‬‬ ‫‪٠.٩٠‬‬ ‫‪٠.١٠‬‬ ‫‪٨‬‬ ‫‪٠.٨١‬‬ ‫‪٠.١٩‬‬ ‫‪٢٨‬‬ ‫‪٠.٥١‬‬ ‫‪٠.٤٩‬‬
‫‪٠.٨٥‬‬ ‫‪٠.١٤‬‬
‫‪٢‬‬ ‫‪٠.٩٥‬‬ ‫‪٠.٠٥‬‬ ‫‪٤‬‬ ‫‪٠.٩٠‬‬ ‫‪٠.١٠‬‬ ‫‪٩‬‬ ‫‪٠.٧٩‬‬ ‫‪٠.٢١‬‬

‫ﺻﻔﺤﺔ ‪228‬‬ ‫ﻣﺟﻠﺔ ﺍﻟﺩﻧﺎﻧﻳﺭ ) ﺍﻟﻌﺩ‪(٢٤‬‬

 ‫ﺍﺳﺘﺨﺪﺍﻡ ﻁﺮﻳﻘﺔ ﺍﻟﻠﻤﺮﺑﻌﺎﺕ ﺍﻟﺼﻐﺮﻯ ﺍﻟﻤﻮﺯﻭﻧﺔ ﻟﺘﻘﺪﻳﺮ ﺍﻟﺒﻘﺎء ﻟﻠﻨﻮﺫﺝ ﺍﻻﺣﺘﻤﺎﻟﻲ ﺍﻟﻤﺨﺘﻠﻂ )ﺍﻻﺳﻲ –ﻛﺎﻣﺎ ﻣﻦ ﺍﻟﺮﺗﺒﺔ ﺍﻟﺜﺎﻧﻴﺔ(‬

‫ﺍﻟﺠﺪﻭﻝ ﺍَﻧﻔﺎ ﻳﺤﺘﻮﻱ ﻋﻠﻰ ﻣﺠﻤﻮﻋﺔ ﻣﻦ ﺍﻻﻋﻤﺪﺓ ﺍﻟﻌﻤﻮﺩ ﺍﻻﻭﻝ )‪ (t‬ﻳﻤﺜﻞ ﺍﻟﺰﻣﻦ ﻭﺍﻟﻌﻤﻮﺩ ﺍﻟﺜﺎﻧﻲ ﻳﻤﺜﻞ ﻗﻴﻢ ﺩﺍﻟﺔ ﺍﻟﺒﻘﺎء ﻭﺍﻟﺘﻲ ﻗﻴﻤﻬﺎ ﺗﺘﻨﺎﻗﺺ‬
‫ﻣﻊ ﺍﻟﺰﻣﻦ " ﺍﻱ ﺗﺘﻨﺎﺳﺐ ﻋﻜﺴﻴﺎ ً ﻣﻊ ﺍﻟﺰﻣﻦ " ‪.‬‬

‫ﺍﻟﻌﻤﻮﺩ ﺍﻟﺜﺎﻟﺚ ﺧﺎﺹ ﺑﺪﺍﻟﺔ ﺍﻟﻜﺜﺎﻓﺔ ﺍﻟﺘﺠﻤﻴﻌﻴﺔ )ﺍﻟﻼﺑﻘﺎء( ﻭﺗﻘﻊ ﻗﻴﻢ ﺍﻟﺪﺍﻟﺔ ﺑﻴﻦ ﺍﻟﺼﻔﺮ"‪ "٠‬ﻭﺍﻟﻮﺍﺣﺪ ﺍﻟﺼﺤﻴﺢ "‪ ،"١‬ﻭﻫﻲ ﻓﻲ ﺗﺰﺍﻳﺪ "‬
‫ﺗﺘﻨﺎﺳﺐ ﻁﺮﺩﻳﺎ ﻣﻊ ﺍﻟﺰﻣﻦ"‪ ،‬ﻭﺍﻥ ﻣﺘﻮﺳﻂ ﻗﻴﻢ ﺩﺍﻟﺔ ﺍﻟﻼﺑﻘﺎء ﻫﻮ )‪ (٠.١٤٣٧٣٦‬ﻭﺑﻨﺴﺒﺔ ﻋﺪﻡ ﺑﻘﺎء ﻋﻠﻰ ﻗﻴﺪ ﺍﻟﺤﻴﺎﺓ ﺑﺎﻟﻨﺴﺒﺔ ﻟﻠﻜﺎﺋﻦ ﺍﻟﺤﻲ‬
‫ﻭﻟﻜﻞ ﺳﺒﻌﺔ ﺍﻳﺎﻡ ﻫﻮ )‪.(٪١٤.٧٣‬‬

‫)‪(Conclusion‬‬ ‫‪ .٥‬ﺍﻻﺳﺘﻨﺘﺎﺟﺎﺕ‬

‫𝑤 ∑( ﻭﺫﻟﻚ ﺑﺄﺩﺧﺎﻟﻬﺎ ﻋﻠﻰ ﺍﻟﺘﻮﺯﻳﻌﺎﺕ ﺍﻟﻤﻔﺮﺩﺓ ﻭﻫﻤﺎ ﺍﻟﺘﻮﺯﻳﻊ )‪(Exponential‬‬ ‫‪ (١‬ﻗﺎﻡ ﺍﻟﺒﺎﺣﺚ ﺑﺄﺳﺘﻌﻤﺎﻝ ﻣﻌﻠﻤﺔ ﺍﻟﺨﻠﻂ )‬
‫‪١‬‬

‫‪ n‬ﻣﻤﺎﻧﺘﺞ ﻋﻨﻪ‬ ‫‪ θ‬ﻭﺷﻜﻞ ‪٣‬‬ ‫‪ θ‬ﻭﺗﻮﺯﻳﻊ )‪ (gamma of the second order‬ﺑﻤﻌﻠﻤﺘﻴﻦ ﻗﻴﺎﺱ ‪٠‬‬ ‫ﺑﻤﻌﻠﻤﺔ ﻗﻴﺎﺱ ‪٠‬‬
‫ﺍﻟﺘﻮﺯﻳﻊ ﺍﻟﻤﺨﺘﻠﻂ )ﺍﻻﺳﻲ– ﻛﺎﻣﺎ ﻣﻦ ﺍﻟﺮﺗﺒﺔ ﺍﻟﺜﺎﻧﻴﺔ (‬
‫‪ (٢‬ﻣﻦ ﺧﻼﻝ ﻣﻌﺎﻳﻴﺮ ﺍﻟﻤﻔﺎﺿﻠﺔ ﺑﻴﻦ ﺍﻟﺘﻮﺯﻳﻌﺎﺕ ﺍﻟﻤﺴﺘﺨﺪﻣﺔ ﻓﻲ ﺍﻟﺠﺎﻧﺐ ﺍﻟﺘﻄﺒﻴﻘﻲ ﻭﺟﺪ ﺍﻟﺒﺎﺣﺚ ﺍﻓﻀﻠﻴﺔ ﺍﻟﺘﻮﺯﻳﻊ ﺍﻟﻤﺨﺘﻠﻂ ﻋﻠﻰ‬
‫ﺍﻟﺘﻮﺯﻳﻌﺎﺕ ﺍﻟﻤﻔﺮﺩﺓ )‪ (Exponentail‬ﻭ )‪ (gamma of the second order‬ﻓﻲ ﺗﻤﺜﻴﻞ ﺍﻟﺒﻴﺎﻧﺎﺕ ﺍﻟﺤﻘﻴﻘﻴﺔ ‪.‬‬

‫)‪(Recommendations‬‬ ‫‪ .٦‬ﺍﻟﺘﻮﺻﻴﺎﺕ‬

‫ﻋﺎﻟﻴﺔ ﻓﻲ ﺗﻤﺜﻴﻞ‬ ‫‪ (١‬ﺍﺳﺘﻌﻤﺎﻝ ﺍﻧﻮﺍﻉ ﺟﺪﻳﺪﺓ ﻣﻦ ﺍﻟﺘﻮﺯﻳﻌﺎﺕ ﺍﻟﻤﺨﺘﻠﻄﺔ ﻭﺫﻟﻚ ﻟﻤﺎ ﺗﻤﺘﺎﺯ ﺑﻪ ﻫﺬﻩ ﺍﻟﺘﻮﺯﻳﻌﺎﺕ ﻣﻦ ﻛﻔﺎءﺓ ﻭﻣﺮﻭﻧﺔ‬
‫ﺑﻴﺎﻧﺎﺕ ﺍﻟﻔﺸﻞ ‪.‬‬
‫‪ (٢‬ﺃﺳﺘﻌﻤﺎﻝ ﻁﺮﺍﺋﻖ ﺗﻘﺪﻳﺮ ﺍﺧﺮﻯ ﻟﺘﻘﺪﻳﺮ ﺩﺍﻟﺔ ﺍﻟﺒﻘﺎء ﻟﻠﺒﻴﺎﻧﺎﺕ ﺍﻟﺤﻘﻴﻘﻴﺔ ﻣﺜﻞ ﻁﺮﻳﻘﺔ ) ﺑﻴﺰ ‪ ،‬ﻭﺍﻳﺖ ‪ ...‬ﺍﻟﺦ ( ‪.‬‬
‫‪ (٣‬ﺍﻋﺘﻤﺎﺩ ﻧﺘﺎﺋﺞ ﺍﻟﺒﺤﺚ ﻣﻦ ﻗﺒﻞ ﻣﺮﻛﺰ ﺍﻣﺮﺍﺽ ﻭﺯﺭﺍﻋﺔ ﺍﻟﻜﻠﻰ – ﻣﺪﻳﻨﺔ ﺍﻟﻄﺐ ‪.‬‬

‫)‪(References‬‬ ‫ﺎدر‬ ‫‪ .٧‬اﻟ‬

‫)‪( Arabic References‬‬ ‫ﺛﺎﻧﻴﺎ ً ‪ :‬ﺍﻟﻤﺼﺎﺩﺭ ﺍﻟﻌﺮﺑﻴﺔ‬

‫‪ -١‬ﺍﻟﺒﻴﺎﺗﻲ ‪ ،‬ﺣﺴﺎﻡ ﻧﺠﻢ ‪) ، (٢٠٠٢) ،‬ﻣﻘﺎﺭﻧﺔ ﻁﺮﺍﺋﻖ ﺗﻘﺪﻳﺮ ﺍﻧﻤﻮﺫﺝ ﻭﻳﺒﻞ ﻟﻠﻔﺸﻞ ﺑﺎﺳﺘﺨﺪﺍﻡ ﺍﻟﻤﺤﺎﻛﺎﺓ ( ‪ ،‬ﺍﻁﺮﻭﺣﺔ ﺩﻛﺘﻮﺭﺍﻩ – ﻛﻠﻴﺔ‬
‫ﺍﻻﺩﺍﺭﺓ ﻭﺍﻻﻗﺘﺼﺎﺩ – ﺟﺎﻣﻌﺔ ﺑﻐﺪﺍﺩ ‪.‬‬

‫‪ -٢‬ﻋﺒﺪ ﷲ ﺍﺣﻤﺪ ﺣﻤﺰﺓ ) ‪ (٢٠١٨‬ﺩﺍﻟﺔ ﺍﻟﺒﻘﺎء ﻟﻤﺮﻳﻀﺎﺕ ﺳﺮﻁﺎﻥ ﺍﻟﺜﺪﻱ ﻓﻲ ﻣﻨﻄﻘﺔ ﺍﻟﻔﺮﺍﺕ ﺍﻻﻭﺳﻂ ‪ ،‬ﺭﺳﺎﻟﺔ ﻣﺎﺟﺴﺘﻴﺮ ‪ -‬ﻛﻠﻴﺔ ﻋﻠﻮﻡ‬
‫ﺍﻟﺤﺎﺳﻮﺏ ﻭﺍﻟﺮﻳﺎﺿﻴﺎﺕ– ﺟﺎﻣﻌﺔ ﺍﻟﻘﺎﺩﺳﻴﺔ‪.‬‬

‫‪ -٣‬ﻣﻄﺮ‪ ،‬ﻅﺎﻓﺮ ﺭﻣﻀﺎﻥ ‪ ،‬ﻭﺍﺧﺮﻭﻥ )‪ (٢٠١١‬ﺩﺭﺍﺳﺔ ﻣﻘﺎﺭﻧﺔ ﻛﻔﺎءﺓﻋﺪﺩ ﻣﻦ ﻣﻌﺎﻳﻴﺮﺍﻟﻤﻌﻠﻮﻣﺎﺕ ﻓﻲ ﺍﺧﺘﻴﺎﺭﻧﻤﺎﺫﺝ ﺍﻟﺴﻼﺳﻞ ﺍﻟﺰﻣﻨﻴﺔ ﻣﻦ ﺍﻟﺮﺗﺐ‬
‫ﺍﻟﺪﻧﻴﺎ – ﺍﻟﻤﺠﻠﺔ ﺍﻟﻌﺮﺍﻗﻴﺔ ﻟﻠﻌﻠﻮﻡ ﺍﻻﺣﺼﺎﺋﻴﺔ ‪. (١٠٠-٧١)٢٢‬‬

‫)‪(Foreign References‬‬ ‫ﺛﺎﻧﻳﺎ‪ :‬ﺍﻟﻣﺻﺎﺩﺭ ﺍﻻﺟﻧﺑﻳﺔ‬

‫ﺻﻔﺤﺔ ‪229‬‬ ‫ﻣﺟﻠﺔ ﺍﻟﺩﻧﺎﻧﻳﺭ ) ﺍﻟﻌﺩ‪(٢٤‬‬

 (‫ﺍﺳﺘﺨﺪﺍﻡ ﻁﺮﻳﻘﺔ ﺍﻟﻠﻤﺮﺑﻌﺎﺕ ﺍﻟﺼﻐﺮﻯ ﺍﻟﻤﻮﺯﻭﻧﺔ ﻟﺘﻘﺪﻳﺮ ﺍﻟﺒﻘﺎء ﻟﻠﻨﻮﺫﺝ ﺍﻻﺣﺘﻤﺎﻟﻲ ﺍﻟﻤﺨﺘﻠﻂ )ﺍﻻﺳﻲ –ﻛﺎﻣﺎ ﻣﻦ ﺍﻟﺮﺗﺒﺔ ﺍﻟﺜﺎﻧﻴﺔ‬

٤. Bakouch ,H.S., Al-Zahrani, B.M., Ali A. Al-Shomrani , Vitor A.A. Marchi,F., Louzada,(٢٠١٢)
,"An extended Lindley distribution " Journal of the Korean Statistical Society ٤١ ٧٥–٨٥.
٥. Chi YY, Ibrahim JG.,٢٠٠٦, " Joint models for multivariate longitudinal and multivariate survival
data.", Biometrics.٦٢(٢):٤٣٢-٤٥.
٦. Cox ,C.;(٢٠٠٨);"The generalized F distribution:an umbrella for parametric survival analysis";
Statistics in Medicine ,vol.٢٧,pp.٤٣٠١-٤٣١٢, U.S.A.
٧. Diab L.S. and , Muhammed H. Z. , (٢٠١٤) "Quasi Lindley Geometric Distribution "
International Journal of Computer Applications (٠٩٧٥ – ٨٨٨٧) Volume ٩٥– No. ١٣, June.
٨. Dolati, A. and H.,Zakerzadeh, (٢٠١٠). Generalized Lindley distribution. Journal of
Mathematical ,Vol.٣,No.٢(٢٠٠٩),١٣-٢٥.
٩. Gupta, R.K., and Kundu, D., (٢٠٠٠), Generalized Exponential Distribution: different method of
estimations, Journal of statistical computation and simulation, vol.٠٠, pp ١-٢٢.
١٠. JelleGoeman , July - ٢٠٠١ , Using survival To predict survival ,Leiden University Medical
Center.
١١. Merovci, F., &Elbatal, I. (٢٠١٤). Transmuted Lindley-geometric distribution and its
applications. Journal of Statistics Applications. Pro. ٣,No. ١, ٧٧-٩١٤٠- Swain, J., Venkatraman,
S. and Wilson, J. (١٩٨٨), “Least Squares Estimation of Distribution Function in Johnson’s
Translation System”,Journal of Statistical Computation and Simulation, ٢٩, ٢٧١-٢٩٧.
١٢. Rinne.H,(٢٠١٤),The Hazard Rate Theory and Inference http://geb.uni-
giessen.de/geb/volltexte/٢٠١٤/١٠٧٩٣/ pdf/RinneHorst_hazardrate_٢٠١٤.pdf.

230 ‫ﺻﻔﺤﺔ‬ (٢٤‫ﻣﺟﻠﺔ ﺍﻟﺩﻧﺎﻧﻳﺭ ) ﺍﻟﻌﺩ‬