Bu t\u00fcr bir grafi\u011fin istatistiksel programlama dili R’de nas\u0131l g\u00f6r\u00fcnd\u00fc\u011f\u00fc a\u015fa\u011f\u0131da a\u00e7\u0131klanm\u0131\u015ft\u0131r:<\/span> <\/p>\n Veri k\u00fcmesindeki her g\u00f6zlem, \u00e7izimde tek bir nokta olarak g\u00f6r\u00fcnt\u00fclenir. X ekseni her noktan\u0131n kald\u0131rac\u0131n\u0131, y ekseni ise her noktan\u0131n standartla\u015ft\u0131r\u0131lm\u0131\u015f kal\u0131nt\u0131s\u0131n\u0131 g\u00f6sterir.<\/span><\/p>\n Kald\u0131ra\u00e7<\/strong> , belirli bir g\u00f6zlemin veri setinden \u00e7\u0131kar\u0131lmas\u0131 durumunda regresyon modelinin katsay\u0131lar\u0131n\u0131n ne \u00f6l\u00e7\u00fcde de\u011fi\u015fece\u011fini ifade eder.<\/span><\/p>\n Y\u00fcksek kald\u0131raca sahip g\u00f6zlemler, regresyon modelinin katsay\u0131lar\u0131 \u00fczerinde g\u00fc\u00e7l\u00fc bir etkiye sahiptir. Bu g\u00f6zlemleri \u00e7\u0131kar\u0131rsak model katsay\u0131lar\u0131 \u00f6nemli \u00f6l\u00e7\u00fcde de\u011fi\u015fecektir.<\/span><\/p>\n Standartla\u015ft\u0131r\u0131lm\u0131\u015f art\u0131klar,<\/strong> bir g\u00f6zlem i\u00e7in tahmin edilen de\u011fer ile g\u00f6zlemin ger\u00e7ek de\u011feri aras\u0131ndaki standartla\u015ft\u0131r\u0131lm\u0131\u015f fark\u0131 ifade eder.<\/span><\/p>\n Bir g\u00f6zlemin standartla\u015ft\u0131r\u0131lm\u0131\u015f bir kal\u0131nt\u0131 i\u00e7in y\u00fcksek bir mutlak de\u011fere sahip olabilece\u011fi, ancak kald\u0131ra\u00e7 i\u00e7in d\u00fc\u015f\u00fck bir de\u011fere sahip olabilece\u011fi dikkate al\u0131nmal\u0131d\u0131r.<\/span><\/p>\n Bu grafikteki bir nokta Cook mesafesinin (k\u0131rm\u0131z\u0131 noktal\u0131 \u00e7izgiler) d\u0131\u015f\u0131nda kal\u0131yorsa, bu etkili bir g\u00f6zlem olarak kabul edilir.<\/span><\/p>\n Daha \u00f6nce g\u00f6sterilen art\u0131klar ve kald\u0131ra\u00e7 grafi\u011fine bakal\u0131m:<\/span> <\/p>\n Yukar\u0131daki \u00f6rnekte 10 numaral\u0131 g\u00f6zlemin Pi\u015firme mesafesi s\u0131n\u0131r\u0131na en yak\u0131n oldu\u011funu ancak noktal\u0131 \u00e7izginin d\u0131\u015f\u0131na \u00e7\u0131kmad\u0131\u011f\u0131n\u0131 g\u00f6rebiliriz.<\/span> Bu, regresyon modelimizde etkili noktalar\u0131n olmad\u0131\u011f\u0131<\/strong> anlam\u0131na gelir.<\/span><\/p>\n Ancak, a\u015fa\u011f\u0131daki art\u0131k\/kald\u0131ra\u00e7 grafi\u011fine sahip oldu\u011fumuzu varsayal\u0131m:<\/span> <\/p>\n Sa\u011f \u00fcst k\u00f6\u015fedeki 1 numaral\u0131 g\u00f6zlemin k\u0131rm\u0131z\u0131 noktal\u0131 \u00e7izgilerin d\u0131\u015f\u0131nda oldu\u011funu g\u00f6rebiliyoruz. Bu da onun bir etki noktas\u0131 oldu\u011funu<\/strong> g\u00f6sterir.<\/span><\/p>\n Bu, e\u011fer bu g\u00f6zlemi veri setimizden \u00e7\u0131kar\u0131rsak ve regresyon modelini tekrar yerle\u015ftirirsek, model katsay\u0131lar\u0131n\u0131n \u00f6nemli \u00f6l\u00e7\u00fcde de\u011fi\u015fece\u011fi anlam\u0131na gelir.<\/span><\/p>\n Bir model i\u00e7in art\u0131klar\u0131n ve kald\u0131rac\u0131n grafi\u011fini olu\u015fturursan\u0131z ve bir veya daha fazla g\u00f6zlemin etkili olarak tan\u0131mland\u0131\u011f\u0131n\u0131 g\u00f6r\u00fcrseniz, birka\u00e7 \u015fey yapabilirsiniz:<\/span><\/p>\n 1. G\u00f6zlemin bir hata olmad\u0131\u011f\u0131n\u0131 do\u011frulay\u0131n.<\/strong><\/span><\/p>\n Herhangi bir eylemde bulunmadan \u00f6nce, \u00f6ncelikle etkili g\u00f6zlemlerin bir veri giri\u015fi hatas\u0131 veya ba\u015fka bir garip olay\u0131n sonucu olmad\u0131\u011f\u0131n\u0131 do\u011frulamal\u0131s\u0131n\u0131z.<\/span><\/p>\n 2. Ba\u015fka bir regresyon modeli uydurmay\u0131 deneyin.<\/strong><\/span><\/p>\n Etkili g\u00f6zlemler, belirtti\u011finiz modelin verilere iyi uymad\u0131\u011f\u0131n\u0131 g\u00f6sterebilir. Bu durumda polinom regresyon modelini<\/a> veya do\u011frusal olmayan modeli deneyebilirsiniz.<\/span><\/p>\n 3. Etkili yorumlar\u0131 kald\u0131r\u0131n.<\/strong><\/span><\/p>\n Son olarak, belirtti\u011finiz model, bir veya iki etkili g\u00f6zlem d\u0131\u015f\u0131nda verilere iyi uyuyorsa, etkili g\u00f6zlemleri kald\u0131rmaya karar verebilirsiniz.<\/span><\/p>\n A\u015fa\u011f\u0131daki e\u011fitimler, regresyon modellerinin uyumunu de\u011ferlendirmek i\u00e7in art\u0131klar\u0131n nas\u0131l kullan\u0131laca\u011f\u0131 hakk\u0131nda ek bilgi sa\u011flar.<\/span><\/p>\n \u0130statistiklerde art\u0131klar nelerdir?<\/a> Art\u0131klar\u0131n kald\u0131raca kar\u015f\u0131 grafi\u011fi, bir regresyon modelinde etkili g\u00f6zlemleri tan\u0131mlamam\u0131za olanak tan\u0131yan bir t\u00fcr te\u015fhis tablosudur . Bu t\u00fcr bir grafi\u011fin istatistiksel programlama dili R’de nas\u0131l g\u00f6r\u00fcnd\u00fc\u011f\u00fc a\u015fa\u011f\u0131da a\u00e7\u0131klanm\u0131\u015ft\u0131r: Veri k\u00fcmesindeki her g\u00f6zlem, \u00e7izimde tek bir nokta olarak g\u00f6r\u00fcnt\u00fclenir. X ekseni her noktan\u0131n kald\u0131rac\u0131n\u0131, y ekseni ise her noktan\u0131n standartla\u015ft\u0131r\u0131lm\u0131\u015f kal\u0131nt\u0131s\u0131n\u0131 g\u00f6sterir. Kald\u0131ra\u00e7 , belirli […]<\/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-2155","post","type-post","status-publish","format-standard","hentry","category-rehber"],"yoast_head":"\n![\"\"](\"https:\/\/statorials.org\/wp-content\/uploads\/2023\/08\/diagnostic2.png\")
<\/p>\n Art\u0131klar ve Kald\u0131ra\u00e7 Grafi\u011fi Nas\u0131l Yorumlan\u0131r?<\/strong><\/span><\/h3>\n
<\/p>\n![\"\"](\"https:\/\/statorials.org\/wp-content\/uploads\/2023\/08\/niveau-1.png\")
<\/p>\n Etkili g\u00f6zlemler nas\u0131l ele al\u0131n\u0131r?<\/strong><\/span><\/h3>\n
Ek kaynaklar<\/strong><\/span><\/h3>\n
Standartla\u015ft\u0131r\u0131lm\u0131\u015f art\u0131klar nelerdir?<\/a>
R’de te\u015fhis grafikleri nas\u0131l yorumlan\u0131r<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"