SAS’ta PCA ger\u00e7ekle\u015ftirmenin en basit yolu, a\u015fa\u011f\u0131daki temel s\u00f6zdizimini kullanan PROC PRINCOMP<\/strong> deyimini kullanmakt\u0131r:<\/span><\/p>\n \u0130\u015fte her talimat\u0131n yapt\u0131\u011f\u0131 \u015fey:<\/span><\/p>\n A\u015fa\u011f\u0131daki ad\u0131m ad\u0131m \u00f6rnek, SAS’ta temel bile\u015fenler analizi ger\u00e7ekle\u015ftirmek i\u00e7in PROC PRINCOMP<\/strong> ifadesinin pratikte nas\u0131l kullan\u0131laca\u011f\u0131n\u0131 g\u00f6sterir.<\/span><\/p>\n 20 basketbol oyuncusu hakk\u0131nda \u00e7e\u015fitli bilgiler i\u00e7eren a\u015fa\u011f\u0131daki veri setine sahip oldu\u011fumuzu varsayal\u0131m:<\/span> <\/p>\n Veri k\u00fcmesinin noktalar\u0131<\/strong> , asistleri<\/strong> ve s\u0131\u00e7rama<\/strong> de\u011fi\u015fkenlerini kullanarak temel bile\u015fen analizini ger\u00e7ekle\u015ftirmek i\u00e7in PROC PRINCOMP<\/strong> ifadesini kullanabiliriz:<\/span><\/span><\/p>\n \u00c7\u0131kt\u0131n\u0131n ilk k\u0131sm\u0131, her bir girdi de\u011fi\u015fkeninin ortalama ve standart sapmalar\u0131, bir korelasyon matrisi ve \u00f6zde\u011ferlerin ve \u00f6zvekt\u00f6rlerin de\u011ferleri dahil olmak \u00fczere \u00e7e\u015fitli tan\u0131mlay\u0131c\u0131 istatistikleri g\u00f6r\u00fcnt\u00fcler:<\/span> <\/p>\n \u00c7\u0131kt\u0131n\u0131n bir sonraki b\u00f6l\u00fcm\u00fcnde bir yama\u00e7 grafi\u011fi<\/strong> ve a\u00e7\u0131klanan varyans<\/strong> grafi\u011fi g\u00f6r\u00fcnt\u00fclenir:<\/span> <\/p>\n PCA uygulad\u0131\u011f\u0131m\u0131zda genellikle veri setindeki toplam varyasyonun y\u00fczde ka\u00e7\u0131n\u0131n her bir temel bile\u015fen taraf\u0131ndan a\u00e7\u0131klanabilece\u011fini anlamak isteriz.<\/span><\/p>\n Ortaya \u00e7\u0131kan Korelasyon Matrisi \u00d6zde\u011ferleri<\/strong> ba\u015fl\u0131kl\u0131 tablo, her bir temel bile\u015fen taraf\u0131ndan toplam varyasyonun y\u00fczde ka\u00e7\u0131n\u0131n a\u00e7\u0131kland\u0131\u011f\u0131n\u0131 tam olarak g\u00f6rmemizi sa\u011flar:<\/span><\/p>\n T\u00fcm y\u00fczdelerin toplam\u0131n\u0131n %100 oldu\u011funu unutmay\u0131n.<\/span><\/p>\n A\u00e7\u0131klanan Varyans<\/strong> ba\u015fl\u0131kl\u0131 \u00e7izim bu de\u011ferleri g\u00f6rselle\u015ftirmemize olanak tan\u0131r.<\/span><\/p>\n X ekseni temel bile\u015feni g\u00f6r\u00fcnt\u00fcler ve y ekseni, her bir temel bile\u015fen taraf\u0131ndan a\u00e7\u0131klanan toplam varyans\u0131n y\u00fczdesini g\u00f6r\u00fcnt\u00fcler.<\/span><\/p>\n Belirli bir veri k\u00fcmesi i\u00e7in PCA sonu\u00e7lar\u0131n\u0131 g\u00f6rselle\u015ftirmek i\u00e7in, bir veri k\u00fcmesindeki her g\u00f6zlemi ilk iki ana bile\u015fen taraf\u0131ndan olu\u015fturulan bir d\u00fczlemde g\u00f6r\u00fcnt\u00fcleyen bir \u00e7izim olan bir biplot<\/strong> olu\u015fturabiliriz.<\/span><\/p>\n Bir biplot olu\u015fturmak i\u00e7in SAS’ta a\u015fa\u011f\u0131daki s\u00f6zdizimini kullanabiliriz:<\/span> <\/p>\n X ekseni birinci ana bile\u015feni, y ekseni ikinci ana bile\u015feni g\u00f6r\u00fcnt\u00fcler ve veri k\u00fcmesindeki bireysel g\u00f6zlemler<\/a> grafi\u011fin i\u00e7inde k\u00fc\u00e7\u00fck daireler halinde g\u00f6r\u00fcnt\u00fclenir.<\/span><\/p>\n Grafikte yan yana yer alan g\u00f6zlemler say\u0131<\/strong> , asist<\/strong> ve ribaund<\/strong> olmak \u00fczere \u00fc\u00e7 de\u011fi\u015fken i\u00e7in de benzer de\u011ferlere sahip.<\/span><\/p>\n \u00d6rne\u011fin grafi\u011fin en solunda 9.<\/strong> ve 10.<\/strong> g\u00f6zlemlerin birbirine son derece yak\u0131n oldu\u011funu g\u00f6rebiliyoruz.<\/span><\/p>\n Orijinal veri setine ba\u015fvurursak bu g\u00f6zlemlere ili\u015fkin \u015fu de\u011ferleri g\u00f6rebiliriz:<\/span><\/p>\n \u00dc\u00e7 de\u011fi\u015fkenin her biri i\u00e7in de\u011ferler benzerdir, bu da bu g\u00f6zlemlerin neden biplotta birbirine bu kadar yak\u0131n oldu\u011funu a\u00e7\u0131klar.<\/span><\/p>\n Korelasyon Matrisi \u00d6zde\u011ferleri<\/strong> ba\u015fl\u0131kl\u0131 sonu\u00e7 tablosunda da ilk iki temel bile\u015fenin veri setindeki toplam varyasyonun %88,21’ini<\/strong> olu\u015fturdu\u011funu g\u00f6rd\u00fck.<\/span><\/p>\n Bu y\u00fczde \u00e7ok y\u00fcksek oldu\u011fundan, iki noktadaki hangi g\u00f6zlemlerin birbirine yak\u0131n oldu\u011funu analiz etmek ge\u00e7erlidir \u00e7\u00fcnk\u00fc iki noktay\u0131 olu\u015fturan iki temel bile\u015fen, veri k\u00fcmesindeki varyasyonun neredeyse tamam\u0131n\u0131 a\u00e7\u0131klamaktad\u0131r.<\/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 basit do\u011frusal regresyon nas\u0131l ger\u00e7ekle\u015ftirilir?<\/a> Temel bile\u015fen analizi (PCA), bir veri k\u00fcmesindeki varyasyonun b\u00fcy\u00fck bir b\u00f6l\u00fcm\u00fcn\u00fc a\u00e7\u0131klayan temel bile\u015fenleri (tahmin edici de\u011fi\u015fkenlerin do\u011frusal kombinasyonlar\u0131) bulmay\u0131 ama\u00e7layandenetimsiz bir makine \u00f6\u011frenme tekni\u011fidir. SAS’ta PCA ger\u00e7ekle\u015ftirmenin en basit yolu, a\u015fa\u011f\u0131daki temel s\u00f6zdizimini kullanan PROC PRINCOMP deyimini kullanmakt\u0131r: proc princomp data =my_data out =out_data outstat =stats; var var1 var2 var3; run ; \u0130\u015fte her […]<\/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-4581","post","type-post","status-publish","format-standard","hentry","category-rehber"],"yoast_head":"\n proc princomp<\/span> data<\/span> =my_data out<\/span> =out_data outstat<\/span> =stats;\n var<\/span> var1 var2 var3;\nrun<\/span> ;<\/strong><\/pre>\n\n
1. Ad\u0131m: Veri k\u00fcmesi olu\u015fturun<\/strong><\/span><\/h2>\n
\/*create dataset*\/\n<\/span>data<\/span> my_data;\n input<\/span> points assists rebounds;\n datalines<\/span> ;\n22 8 4\n29 7 3\n10 4 12\n5 5 15\n35 6 2\n8 3 10\n10 4 8\n8 4 3\n2 5 17\n4 5 19\n9 9 4\n7 6 4\n31 5 3\n4 6 13\n5 7 8\n8 8 4\n10 4 8\n20 4 6\n25 8 8\n18 8 3\n;\nrun<\/span> ;\n\n\/*view dataset*\/\n<\/span>proc print<\/span> data<\/span> =my_data;<\/strong> <\/pre>\n![\"\"](\"https:\/\/statorials.org\/wp-content\/uploads\/2023\/08\/pcasas1.png\")
<\/p>\n Ad\u0131m 2: Temel bile\u015fenler analizini ger\u00e7ekle\u015ftirin<\/strong><\/span><\/h2>\n
\/*perform principal components analysis*\/<\/span>\nproc princomp<\/span> data<\/span> =my_data out<\/span> =out_data outstat<\/span> =stats;\n var<\/span> points assists rebounds;\nrun<\/span> ;<\/strong><\/pre>\n![\"\"](\"https:\/\/statorials.org\/wp-content\/uploads\/2023\/08\/pcasas2.png\")
<\/p>\n![\"\"](\"https:\/\/statorials.org\/wp-content\/uploads\/2023\/08\/pcasas3.png\")
<\/p>\n\n
3. Ad\u0131m: Sonu\u00e7lar\u0131 g\u00f6rselle\u015ftirmek i\u00e7in bir ikili grafik olu\u015fturun<\/strong><\/span><\/h2>\n
\/*create dataset with column called obs<\/em> to represent row numbers of original data*\/<\/span>\ndata<\/span> biplot_data;\n set<\/span> out_data;\n obs=_n_;\nrun<\/span> ;\n\n\/*create biplot using values from first two principal components*\/\n<\/span>proc sgplot<\/span> data<\/span> =biplot_data;\n scatter<\/span> x<\/span> =Prin1 y<\/span> =Prin2 \/ datalabel<\/span> =obs;\nrun<\/span> ;<\/strong><\/span> <\/pre>\n![\"\"](\"https:\/\/statorials.org\/wp-content\/uploads\/2023\/08\/pcasas4.png\")
<\/p>\n\n
Ek kaynaklar<\/strong><\/span><\/h2>\n
SAS’ta \u00e7oklu do\u011frusal regresyon nas\u0131l ger\u00e7ekle\u015ftirilir?<\/a>
SAS’ta lojistik regresyon nas\u0131l ger\u00e7ekle\u015ftirilir?<\/a><\/span><\/p>\n","protected":false},"excerpt":{"rendered":"