Tipik olarak do\u011frusal regresyon uygulad\u0131\u011f\u0131m\u0131zda yan\u0131t de\u011fi\u015fkeninin ortalama de\u011ferini tahmin etmek isteriz.<\/span><\/p>\n Bununla birlikte, yan\u0131t de\u011ferinin herhangi bir<\/em> nicelik veya y\u00fczdelik de\u011ferini (70. y\u00fczdelik dilim, 90. y\u00fczdelik dilim, 98. y\u00fczdelik dilim vb.) tahmin etmek i\u00e7in bunun yerine nicelik regresyonu<\/strong> olarak bilinen bir y\u00f6ntemi kullanabiliriz.<\/span><\/p>\n R’de niceliksel regresyon ger\u00e7ekle\u015ftirmek i\u00e7in, a\u015fa\u011f\u0131daki s\u00f6zdizimini kullanan quantreg<\/a> paketindeki rq()<\/strong> i\u015flevini kullanabiliriz:<\/span><\/p>\n Alt\u0131n:<\/span><\/p>\n Bu e\u011fitimde, R’de niceliksel regresyon ger\u00e7ekle\u015ftirmek i\u00e7in bu i\u015flevin nas\u0131l kullan\u0131laca\u011f\u0131na ili\u015fkin ad\u0131m ad\u0131m bir \u00f6rnek sunulmaktad\u0131r.<\/span><\/p>\n Bu \u00f6rnek i\u00e7in bir \u00fcniversitede 100 farkl\u0131 \u00f6\u011frencinin ders \u00e7al\u0131\u015fma saatlerini ve elde etti\u011fi s\u0131nav sonu\u00e7lar\u0131n\u0131 i\u00e7eren bir veri seti olu\u015fturaca\u011f\u0131z:<\/span><\/p>\n Daha sonra, yorday\u0131c\u0131 de\u011fi\u015fken olarak \u00e7al\u0131\u015f\u0131lan saatleri ve yan\u0131t de\u011fi\u015fkeni olarak s\u0131nav puanlar\u0131n\u0131 kullanarak niceliksel bir regresyon modeli uygulayaca\u011f\u0131z.<\/span><\/p>\n Bu modeli, \u00e7al\u0131\u015f\u0131lan saat say\u0131s\u0131na ba\u011fl\u0131 olarak s\u0131nav puanlar\u0131n\u0131n beklenen y\u00fczde 90’l\u0131k dilimini tahmin etmek i\u00e7in kullanaca\u011f\u0131z:<\/span><\/p>\n Sonu\u00e7tan tahmini regresyon denklemini g\u00f6rebiliriz:<\/span><\/p>\n 90. y\u00fczdelik s\u0131nav puan\u0131 = 60,25 + 2,437*(saat)<\/span><\/p>\n \u00d6rne\u011fin 8 saat e\u011fitim g\u00f6ren t\u00fcm \u00f6\u011frencilerin 90. y\u00fczdelik puan\u0131 79,75 olmal\u0131d\u0131r:<\/span><\/p>\n S\u0131nav puan\u0131n\u0131n 90. y\u00fczdelik dilimi = 60,25 + 2,437*(8) = 79,75<\/strong> .<\/span><\/p>\n \u00c7\u0131kt\u0131 ayr\u0131ca tahmin de\u011fi\u015fkeninin kesi\u015fme noktas\u0131 ve s\u00fcreleri i\u00e7in \u00fcst ve alt g\u00fcven s\u0131n\u0131rlar\u0131n\u0131 da g\u00f6r\u00fcnt\u00fcler.<\/span><\/p>\n Grafi\u011fe yerle\u015ftirilmi\u015f niceliksel regresyon denklemi ile bir da\u011f\u0131l\u0131m grafi\u011fi olu\u015fturarak regresyon sonu\u00e7lar\u0131n\u0131 da g\u00f6rselle\u015ftirebiliriz:<\/span> <\/p>\n Geleneksel do\u011frusal regresyon \u00e7izgisinin aksine, bu uydurulmu\u015f \u00e7izginin verilerin merkezinden ge\u00e7medi\u011fini unutmay\u0131n. Bunun yerine, yorday\u0131c\u0131 de\u011fi\u015fkenin her d\u00fczeyinde tahmini 90. y\u00fczdelik dilimden ge\u00e7er.<\/span><\/p>\n Uydurulmu\u015f kantil regresyon denklemi ile basit do\u011frusal regresyon denklemi aras\u0131ndaki fark\u0131 geom_smooth()<\/strong> arg\u00fcman\u0131n\u0131 ekleyerek g\u00f6rebiliriz:<\/span> <\/p>\n Siyah \u00e7izgi, 90’\u0131nc\u0131 y\u00fczdelik i\u00e7in ayarlanan kantil regresyon \u00e7izgisini, mavi \u00e7izgi ise yan\u0131t de\u011fi\u015fkeninin ortalama de\u011ferini tahmin eden basit do\u011frusal regresyon \u00e7izgisini g\u00f6sterir.<\/span><\/p>\n Beklendi\u011fi gibi, basit do\u011frusal regresyon \u00e7izgisi veriler \u00fczerinden ge\u00e7iyor ve bize her saat d\u00fczeyinde s\u0131nav puanlar\u0131n\u0131n tahmini ortalama de\u011ferini g\u00f6steriyor.<\/span><\/p>\n A\u015fa\u011f\u0131daki e\u011fitimlerde R’de di\u011fer ortak g\u00f6revlerin nas\u0131l ger\u00e7ekle\u015ftirilece\u011fi a\u00e7\u0131klanmaktad\u0131r:<\/span><\/p>\n R’de basit do\u011frusal regresyon nas\u0131l ger\u00e7ekle\u015ftirilir<\/a> Do\u011frusal regresyon, bir veya daha fazla yorday\u0131c\u0131 de\u011fi\u015fken ile bir yan\u0131t de\u011fi\u015fkeni aras\u0131ndaki ili\u015fkiyi anlamak i\u00e7in kullanabilece\u011fimiz bir y\u00f6ntemdir. Tipik olarak do\u011frusal regresyon uygulad\u0131\u011f\u0131m\u0131zda yan\u0131t de\u011fi\u015fkeninin ortalama de\u011ferini tahmin etmek isteriz. Bununla birlikte, yan\u0131t de\u011ferinin herhangi bir nicelik veya y\u00fczdelik de\u011ferini (70. y\u00fczdelik dilim, 90. y\u00fczdelik dilim, 98. y\u00fczdelik dilim vb.) tahmin etmek i\u00e7in bunun […]<\/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-1325","post","type-post","status-publish","format-standard","hentry","category-rehber"],"yoast_head":"\n library<\/span> (quantreg)\n\nmodel <- rq(y ~ x, data = dataset, tau = 0.5<\/span> )\n<\/strong><\/pre>\n\n
1. Ad\u0131m: Verileri girin<\/strong><\/span><\/h2>\n
#make this example reproducible\n<\/span>set.seed(0)\n\n#create data frame<\/span> \nhours <- runif(100, 1, 10)\nscore <- 60 + 2*hours + rnorm(100, mean=0, sd=.45*hours)\ndf <- data.frame(hours, score)\n\n#view first six rows<\/span>\nhead(df)\n\n hours score\n1 9.070275 79.22682\n2 3.389578 66.20457\n3 4.349115 73.47623\n4 6.155680 70.10823\n5 9.173870 78.12119\n6 2.815137 65.94716\n<\/strong><\/pre>\n Ad\u0131m 2: Kantil Regresyon Ger\u00e7ekle\u015ftirin<\/strong><\/span><\/h2>\n
library<\/span> (quantreg)\n\n#fit model\n<\/span>model <- rq(score ~ hours, data = df, tau = 0.9<\/span> )\n\n#view summary of model\n<\/span>summary(model)\n\nCall: rq(formula = score ~ hours, tau = 0.9, data = df)\n\ntau: [1] 0.9\n\nCoefficients:\n coefficients lower bd upper bd\n(Intercept) 60.25185 59.27193 62.56459\nhours 2.43746 1.98094 2.76989\n<\/strong><\/pre>\n 3. Ad\u0131m: Sonu\u00e7lar\u0131 g\u00f6rselle\u015ftirin<\/strong><\/span><\/h2>\n
library<\/span> (ggplot2)\n\n#create scatterplot with quantile regression line\n<\/span>ggplot(df, aes(hours,score)) +\n geom_point() + \n geom_abline(intercept= coef<\/span> (model)[1], slope= coef<\/span> (model)[2])<\/strong> <\/pre>\n![\"R'de](\"https:\/\/statorials.org\/wp-content\/uploads\/2023\/08\/quantileregr1.png\")
<\/p>\n library<\/span> (ggplot2)\n\n#create scatterplot with quantile regression line and<\/em> simple linear regression line\n<\/span>ggplot(df, aes(hours,score)) +\n geom_point() + \n geom_abline(intercept= coef<\/span> (model)[1], slope= coef<\/span> (model)[2]) +\n geom_smooth(method=\" lm<\/span> \", se= F<\/span> )<\/strong> <\/pre>\n![\"Kantil](\"https:\/\/statorials.org\/wp-content\/uploads\/2023\/08\/quantileregr2.png\")
<\/p>\n Ek kaynaklar<\/strong><\/span><\/h2>\n
R’de \u00e7oklu do\u011frusal regresyon nas\u0131l ger\u00e7ekle\u015ftirilir<\/a>
R’de ikinci dereceden regresyon nas\u0131l ger\u00e7ekle\u015ftirilir?<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"