{"id":4230,"date":"2023-07-12T16:42:36","date_gmt":"2023-07-12T16:42:36","guid":{"rendered":"https:\/\/statorials.org\/tr\/adim-adim-gerileme\/"},"modified":"2023-07-12T16:42:36","modified_gmt":"2023-07-12T16:42:36","slug":"adim-adim-gerileme","status":"publish","type":"post","link":"https:\/\/statorials.org\/tr\/adim-adim-gerileme\/","title":{"rendered":"Sas'ta a\u015famal\u0131 regresyon nas\u0131l ger\u00e7ekle\u015ftirilir (\u00f6rnekle)"},"content":{"rendered":"

<\/p>\n

\n

Ad\u0131m ad\u0131m regresyon,<\/a> tahmin edicileri modele ad\u0131m ad\u0131m girerek ve \u00e7\u0131kararak bir dizi tahmin de\u011fi\u015fkeninden bir regresyon modeli olu\u015fturmak i\u00e7in kullanabilece\u011fimiz bir prosed\u00fcrd\u00fcr. daha fazlas\u0131n\u0131 sil.<\/span><\/p>\n

Ad\u0131m ad\u0131m regresyonun amac\u0131 , yan\u0131t de\u011fi\u015fkeniyle<\/a> istatistiksel olarak anlaml\u0131 bir \u015fekilde ili\u015fkili olan t\u00fcm yorday\u0131c\u0131 de\u011fi\u015fkenleri i\u00e7eren bir regresyon modeli olu\u015fturmakt\u0131r.<\/span><\/p>\n

SAS’ta ad\u0131m ad\u0131m regresyon ger\u00e7ekle\u015ftirmek i\u00e7in PROC REG’i<\/strong> SELECTION<\/strong> deyimiyle birlikte kullanabilirsiniz.<\/span><\/p>\n

A\u015fa\u011f\u0131daki \u00f6rnek, pratikte SAS’ta ad\u0131m ad\u0131m regresyonun nas\u0131l ger\u00e7ekle\u015ftirilece\u011fini g\u00f6sterir.<\/span><\/p>\n

\u00d6rnek: SAS’ta Ad\u0131m Ad\u0131m Regresyon Ger\u00e7ekle\u015ftirme<\/span><\/strong><\/h2>\n

SAS’ta d\u00f6rt \u00f6ng\u00f6r\u00fcc\u00fc de\u011fi\u015fken (x1, x2, x3, x4) ve bir yan\u0131t de\u011fi\u015fkeni (y) i\u00e7eren a\u015fa\u011f\u0131daki veri k\u00fcmesine sahip oldu\u011fumuzu varsayal\u0131m:<\/span> <\/p>\n

 \/*create dataset*\/\n<\/span>data<\/span> my_data;\n    input<\/span> x1 x2 x3 x4 y;\n    datalines<\/span> ;\n1 4 10 13 78\n2 4 12 14 81\n5 3 7 10 75\n8 2 13 9 97\n10 5 12 5 95\n14 7 8 6 90\n17 8 10 6 86 \n19 5 15 5 90\n20 5 12 4 93\n21 4 10 3 95\n;\nrun<\/span> ;\n\n\/*view dataset*\/\n<\/span>proc print<\/span> data<\/span> =my_data;\n<\/strong><\/pre>\n
\"\"<\/p>\n
 \u201cEn iyi\u201d regresyon modelinden bahsetti\u011fimizde, belirli \u00f6nlemleri maksimuma \u00e7\u0131karan veya minimuma indiren modeli kastediyoruz.<\/span><\/p>\n
 Bir grup potansiyel model aras\u0131nda hangi regresyon modelinin en iyi oldu\u011funu de\u011ferlendirmek i\u00e7in yayg\u0131n olarak kulland\u0131\u011f\u0131m\u0131z iki \u00f6l\u00e7\u00fcm vard\u0131r:<\/span><\/p>\n
 1. D\u00fczeltilmi\u015f R-kare<\/strong> : Ayarlanm\u0131\u015f R-kare de\u011feri,<\/a> bir modeldeki tahminlerin say\u0131s\u0131na g\u00f6re ayarlanan bir modelin kullan\u0131\u015fl\u0131l\u0131\u011f\u0131n\u0131 bize s\u00f6yler. En y\u00fcksek d\u00fczeltilmi\u015f R-kare de\u011ferine sahip model en iyi model olarak kabul edilir.<\/span><\/p>\n
 2. AIC<\/strong> : Akaike Bilgi Kriteri<\/a> (AIC), farkl\u0131 regresyon modellerinin uyumunu kar\u015f\u0131la\u015ft\u0131rmak i\u00e7in kullan\u0131lan bir \u00f6l\u00e7\u00fcmd\u00fcr. En d\u00fc\u015f\u00fck AIC de\u011ferine sahip model en iyi model olarak kabul edilir.<\/span><\/p>\n
 Neyse ki, SAS’taki regresyon modelleri i\u00e7in hem uygun R-kare hem de AIC de\u011ferlerini PROC REG’i<\/strong> SELECTION<\/strong> ifadesi ile kullanarak hesaplayabiliyoruz.<\/span><\/p>\n
 A\u015fa\u011f\u0131daki kod bunun nas\u0131l yap\u0131laca\u011f\u0131n\u0131 g\u00f6sterir:<\/span> <\/p>\n
 \/*perform stepwise multiple linear regression*\/\n<\/span>proc reg<\/span> data<\/span> =my_data outest<\/span> =est;\n    model<\/span> y=x1 x2 x3 x4 \/ selection=adjrsq aic ;\n    output out<\/span> =out p=pr=r;\nrun<\/span> ;\nquit<\/span> ; \n<\/strong><\/pre>\n
\"SAS'ta<\/p>\n
 \u00c7\u0131kt\u0131, olas\u0131 her \u00e7oklu do\u011frusal regresyon modeli i\u00e7in uygun R-kare ve AIC de\u011ferlerini g\u00f6r\u00fcnt\u00fcler.<\/span><\/p>\n
 Sonu\u00e7tan, d\u00fczeltilmi\u015f R kare de\u011feri en y\u00fcksek ve<\/em> AIC de\u011feri en d\u00fc\u015f\u00fck olan de\u011ferin, yorday\u0131c\u0131 de\u011fi\u015fken olarak yaln\u0131zca x3 ve x4’\u00fc kullanan regresyon modeli oldu\u011funu g\u00f6rebiliriz.<\/span><\/p>\n
 Bu nedenle a\u015fa\u011f\u0131daki modelin olas\u0131 t\u00fcm modeller aras\u0131nda “en iyisi” oldu\u011funu beyan ederiz:<\/span><\/p>\n
 y = b 0<\/sub> + b 1<\/sub> (x3) + b 2<\/sub> (x4)<\/span><\/p>\n
 Bu \u00f6zel regresyon modeli a\u015fa\u011f\u0131daki \u00f6l\u00e7\u00fcmlere sahiptir:<\/span><\/p>\n
 \u015eimdi, hangi yorday\u0131c\u0131 de\u011fi\u015fken kombinasyonunun en iyi \u00e7oklu do\u011frusal regresyon modelini<\/a> \u00fcretece\u011fini belirlemek istedi\u011fimizi varsayal\u0131m.<\/span><\/p>\n