Cook mesafesinin form\u00fcl\u00fc:<\/span><\/p>\n d ben<\/sub><\/strong> = (r i<\/sub> 2<\/sup> \/ p*MSE) * (h ii<\/sub> \/ (1-h ii<\/sub> ) 2<\/sup> )<\/span><\/p>\n Alt\u0131n:<\/span><\/p>\n Cook mesafesi esasen i’inci<\/sup> g\u00f6zlem kald\u0131r\u0131ld\u0131\u011f\u0131nda modelin t\u00fcm uyum de\u011ferlerinin ne kadar de\u011fi\u015fti\u011fini \u00f6l\u00e7er.<\/span><\/p>\n Cook mesafesinin de\u011feri ne kadar b\u00fcy\u00fck olursa, belirli bir g\u00f6zlem o kadar etkili olur.<\/span><\/p>\n Genel olarak Cook mesafesi 4\/n’den b\u00fcy\u00fck olan herhangi bir g\u00f6zlemin (burada n<\/em> = toplam g\u00f6zlemler) b\u00fcy\u00fck bir etkiye sahip oldu\u011fu kabul edilir.<\/span><\/p>\n A\u015fa\u011f\u0131daki \u00f6rnek, SAS’taki bir regresyon modelinde her g\u00f6zlem i\u00e7in Cook mesafesinin nas\u0131l hesaplanaca\u011f\u0131n\u0131 g\u00f6sterir.<\/span><\/p>\n SAS’ta a\u015fa\u011f\u0131daki veri setine sahip oldu\u011fumuzu varsayal\u0131m:<\/span> <\/p>\n Bu veri setine basit bir do\u011frusal regresyon modeli<\/a> uydurmak i\u00e7in PROC REG’i<\/strong> kullanabiliriz, ard\u0131ndan regresyon modelindeki her g\u00f6zlem i\u00e7in Cook mesafesini hesaplamak amac\u0131yla OUTPUT<\/strong> ifadesini COOKD<\/strong> ifadesi ile birlikte kullanabiliriz:<\/span><\/p>\n Nihai sonu\u00e7 tablosu, her g\u00f6zlem i\u00e7in orijinal veri k\u00fcmesini Pi\u015firme mesafesiyle birlikte g\u00f6r\u00fcnt\u00fcler:<\/span> <\/p>\n \u00d6rne\u011fin \u015funu g\u00f6rebiliriz:<\/span><\/p>\n Ve benzeri.<\/span><\/p>\n PROC REG<\/strong> prosed\u00fcr\u00fc ayn\u0131 zamanda \u00e7\u0131kt\u0131da birka\u00e7 te\u015fhis grafi\u011fi \u00fcretir ve Pi\u015firme mesafesi grafi\u011fi bu \u00e7\u0131kt\u0131da g\u00f6r\u00fclebilir:<\/span> <\/p>\n X ekseni g\u00f6zlem say\u0131s\u0131n\u0131, y ekseni ise her g\u00f6zlem i\u00e7in Pi\u015firme mesafesini g\u00f6sterir.<\/span><\/p>\n 4\/n’ye bir kesme \u00e7izgisi yerle\u015ftirildi\u011fine dikkat edin (bu durumda n = 12, dolay\u0131s\u0131yla kesme 0,33’t\u00fcr) ve veri k\u00fcmesindeki \u00fc\u00e7 g\u00f6zlemin bu \u00e7izgiden daha b\u00fcy\u00fck oldu\u011funu g\u00f6rebiliriz.<\/span><\/p>\n Bu, bu g\u00f6zlemlerin regresyon modeli \u00fczerinde b\u00fcy\u00fck bir etkiye sahip olabilece\u011fini ve belki de model sonu\u00e7lar\u0131n\u0131 yorumlamadan \u00f6nce daha yak\u0131ndan incelenmesi gerekti\u011fini g\u00f6stermektedir.<\/span><\/p>\n A\u015fa\u011f\u0131daki e\u011fitimlerde SAS’ta di\u011fer ortak g\u00f6revlerin nas\u0131l ger\u00e7ekle\u015ftirilece\u011fi a\u00e7\u0131klanmaktad\u0131r:<\/span><\/p>\n SAS’ta Art\u0131k Grafik Nas\u0131l Olu\u015fturulur<\/a> Cook mesafesi, bir regresyon modelinde etkili g\u00f6zlemleri tan\u0131mlamak i\u00e7in kullan\u0131l\u0131r. Cook mesafesinin form\u00fcl\u00fc: d ben = (r i 2 \/ p*MSE) * (h ii \/ (1-h ii ) 2 ) Alt\u0131n: r i i’inci kal\u0131nt\u0131d\u0131r p , regresyon modelindeki katsay\u0131lar\u0131n say\u0131s\u0131d\u0131r MSE ortalama kare hatas\u0131d\u0131r h ii i’inci kald\u0131ra\u00e7 de\u011feridir Cook mesafesi esasen i’inci g\u00f6zlem […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[11],"tags":[],"class_list":["post-4525","post","type-post","status-publish","format-standard","hentry","category-rehber"],"yoast_head":"\n\n
\u00d6rnek: SAS’ta a\u015f\u00e7\u0131n\u0131n mesafesini hesaplama<\/strong><\/h2>\n
\/*create dataset*\/\n<\/span>data<\/span> my_data;\n input<\/span> xy;\n datalines<\/span> ;\n8 41\n12 42\n12 39\n13 37\n14 35\n16 39\n17 45\n22 46\n24 39\n26 49\n29 55\n30 57\n;\nrun<\/span> ;\n\n\/*view dataset*\/\n<\/span>proc print<\/span> data<\/span> =my_data;\n<\/strong><\/span><\/pre>\n![\"\"](\"https:\/\/statorials.org\/wp-content\/uploads\/2023\/08\/cuisiniers1.png\")
<\/p>\n \/*fit simple linear regression model and calculate Cook's distance for each obs*\/\n<\/span>proc reg<\/span> data<\/span> =my_data;\n model<\/span> y=x;\n output<\/span> out=cooksData cookd<\/span> =cookd;\nrun<\/span> ;\n\n\/*print Cook's distance values for each observation*\/\n<\/span>proc print<\/span> data<\/span> =cooksData;\n<\/span><\/strong><\/span><\/pre>\n![\"\"](\"https:\/\/statorials.org\/wp-content\/uploads\/2023\/08\/cuisiniers2.png\")
<\/p>\n\n
![\"SAS'ta](\"https:\/\/statorials.org\/wp-content\/uploads\/2023\/08\/cuisiniers3.png\")
<\/p>\n Ek kaynaklar<\/strong><\/span><\/h2>\n
SAS’ta Histogramlar Nas\u0131l Olu\u015fturulur<\/a>
SAS’ta nokta bulutlar\u0131 nas\u0131l olu\u015fturulur?<\/a>
SAS’ta ayk\u0131r\u0131 de\u011ferler nas\u0131l belirlenir<\/a><\/span><\/p>\n","protected":false},"excerpt":{"rendered":"