A\u015fa\u011f\u0131daki ad\u0131m ad\u0131m \u00f6rnek, R’de spline regresyonunun nas\u0131l ger\u00e7ekle\u015ftirilece\u011fini g\u00f6sterir.<\/span><\/p>\n \u00d6ncelikle R’de iki de\u011fi\u015fkenli bir veri seti olu\u015ftural\u0131m ve de\u011fi\u015fkenler aras\u0131ndaki ili\u015fkiyi g\u00f6rselle\u015ftirmek i\u00e7in bir da\u011f\u0131l\u0131m grafi\u011fi olu\u015ftural\u0131m:<\/span> <\/p>\n A\u00e7\u0131k\u00e7as\u0131, x ve y aras\u0131ndaki ili\u015fki do\u011frusal de\u011fildir ve verideki modelin x=7 ve x=10’da aniden de\u011fi\u015fti\u011fi iki nokta veya “d\u00fc\u011f\u00fcm” var gibi g\u00f6r\u00fcnmektedir.<\/span><\/p>\n Daha sonra bu veri setine basit bir do\u011frusal regresyon modeli s\u0131\u011fd\u0131rmak i\u00e7in lm() fonksiyonunu<\/a> kullanal\u0131m ve da\u011f\u0131l\u0131m grafi\u011finde regresyon do\u011frusunu uygun \u015fekilde \u00e7izelim:<\/span> <\/p>\n Da\u011f\u0131l\u0131m grafi\u011finden basit do\u011frusal regresyon \u00e7izgisinin verilere iyi uymad\u0131\u011f\u0131n\u0131 g\u00f6rebiliriz.<\/span><\/p>\n Model sonu\u00e7lar\u0131ndan d\u00fczeltilmi\u015f R-kare de\u011ferinin<\/a> 0,5903<\/strong> oldu\u011funu da g\u00f6rebiliyoruz.<\/span><\/p>\n Bunu bir spline modelinin d\u00fczeltilmi\u015f R-kare de\u011feriyle kar\u015f\u0131la\u015ft\u0131raca\u011f\u0131z.<\/span><\/p>\n Daha sonra, iki d\u00fc\u011f\u00fcml\u00fc bir spline regresyon modelini s\u0131\u011fd\u0131rmak i\u00e7in splines<\/strong> paketindeki bs()<\/strong> i\u015flevini kullanal\u0131m ve ard\u0131ndan tak\u0131lan modeli da\u011f\u0131l\u0131m grafi\u011finde \u00e7izelim:<\/span> <\/p>\n Da\u011f\u0131l\u0131m grafi\u011finden spline regresyon modelinin verilere olduk\u00e7a iyi uyum sa\u011flayabildi\u011fini g\u00f6rebiliriz.<\/span><\/p>\n Model sonu\u00e7lar\u0131ndan d\u00fczeltilmi\u015f R-kare de\u011ferinin 0,8987<\/strong> oldu\u011funu da g\u00f6rebiliyoruz.<\/span><\/p>\n Bu model i\u00e7in d\u00fczeltilmi\u015f R-kare de\u011feri, basit do\u011frusal regresyon modelinden \u00e7ok daha y\u00fcksektir, bu da bize spline regresyon modelinin verilere daha iyi uyum sa\u011flayabildi\u011fini s\u00f6yler.<\/span><\/p>\n Bu \u00f6rnekte d\u00fc\u011f\u00fcmlerin x=7 ve x=10’da konumland\u0131\u011f\u0131n\u0131 se\u00e7ti\u011fimize dikkat edin.<\/span><\/p>\n Uygulamada, verilerdeki kal\u0131plar\u0131n de\u011fi\u015fti\u011fi yere ve alan uzmanl\u0131\u011f\u0131n\u0131za ba\u011fl\u0131 olarak d\u00fc\u011f\u00fcm konumlar\u0131n\u0131 kendiniz se\u00e7meniz gerekecektir.<\/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 \u00e7oklu do\u011frusal regresyon nas\u0131l ger\u00e7ekle\u015ftirilir<\/a> Spline regresyon, verilerdeki desenin aniden de\u011fi\u015fti\u011fi noktalar veya “d\u00fc\u011f\u00fcmler” oldu\u011funda ve do\u011frusal regresyon ile polinom regresyonun verilere uyacak kadar esnek olmad\u0131\u011f\u0131 durumlarda kullan\u0131lan bir regresyon t\u00fcr\u00fcd\u00fcr. A\u015fa\u011f\u0131daki ad\u0131m ad\u0131m \u00f6rnek, R’de spline regresyonunun nas\u0131l ger\u00e7ekle\u015ftirilece\u011fini g\u00f6sterir. 1. Ad\u0131m: Verileri olu\u015fturun \u00d6ncelikle R’de iki de\u011fi\u015fkenli bir veri seti olu\u015ftural\u0131m ve de\u011fi\u015fkenler aras\u0131ndaki ili\u015fkiyi g\u00f6rselle\u015ftirmek i\u00e7in bir […]<\/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-3998","post","type-post","status-publish","format-standard","hentry","category-rehber"],"yoast_head":"\n 1. Ad\u0131m: Verileri olu\u015fturun<\/strong><\/span><\/h2>\n
#create data frame\n<\/span>df <- data. frame<\/span> (x=1:20,\n y=c(2, 4, 7, 9, 13, 15, 19, 16, 13, 10,\n 11, 14, 15, 15, 16, 15, 17, 19, 18, 20))\n\n#view head of data frame\n<\/span>head(df)\n\n xy\n1 1 2\n2 2 4\n3 3 7\n4 4 9\n5 5 13\n6 6 15\n\n#create scatterplot\n<\/span>plot(df$x, df$y, cex= 1.5<\/span> , pch= 19<\/span> )\n<\/strong><\/pre>\n![\"\"](\"https:\/\/statorials.org\/wp-content\/uploads\/2023\/08\/spline1.jpg\")
<\/p>\n Ad\u0131m 2: Basit do\u011frusal regresyon modelini yerle\u015ftirin<\/strong><\/span><\/h2>\n
#fit simple linear regression model\n<\/span>linear_fit <- lm(df$y ~ df$x)\n\n#view model summary\n<\/span>summary(linear_fit)\n\nCall:\nlm(formula = df$y ~ df$x)\n\nResiduals:\n Min 1Q Median 3Q Max \n-5.2143 -1.6327 -0.3534 0.6117 7.8789 \n\nCoefficients:\n Estimate Std. Error t value Pr(>|t|) \n(Intercept) 6.5632 1.4643 4.482 0.000288 ***\ndf$x 0.6511 0.1222 5.327 4.6e-05 ***\n---\nSignificant. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1\n\nResidual standard error: 3.152 on 18 degrees of freedom\nMultiple R-squared: 0.6118, Adjusted R-squared: 0.5903 \nF-statistic: 28.37 on 1 and 18 DF, p-value: 4.603e-05\n\n#create scatterplot\n<\/span>plot(df$x, df$y, cex= 1.5<\/span> , pch= 19<\/span> )\n\n#add regression line to scatterplot\n<\/span>abline(linear_fit)<\/strong> <\/pre>\n![\"\"](\"https:\/\/statorials.org\/wp-content\/uploads\/2023\/08\/spline2.jpg\")
<\/p>\n Ad\u0131m 3: Spline regresyon modelini yerle\u015ftirin<\/strong><\/span><\/h2>\n
library<\/span> (splines)\n\n#fit spline regression model\n<\/span>spline_fit <- lm(df$y ~ bs(df$x, knots=c( 7<\/span> , 10<\/span> )))\n\n#view summary of spline regression model\n<\/span>summary(spline_fit)\n\nCall:\nlm(formula = df$y ~ bs(df$x, knots = c(7, 10)))\n\nResiduals:\n Min 1Q Median 3Q Max \n-2.84883 -0.94928 0.08675 0.78069 2.61073 \n\nCoefficients:\n Estimate Std. Error t value Pr(>|t|) \n(Intercept) 2.073 1.451 1.429 0.175 \nbs(df$x, knots = c(7, 10))1 2.173 3.247 0.669 0.514 \nbs(df$x, knots = c(7, 10))2 19.737 2.205 8.949 3.63e-07 ***\nbs(df$x, knots = c(7, 10))3 3.256 2.861 1.138 0.274 \nbs(df$x, knots = c(7, 10))4 19.157 2.690 7.121 5.16e-06 ***\nbs(df$x, knots = c(7, 10))5 16.771 1.999 8.391 7.83e-07 ***\n---\nSignificant. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1\n\nResidual standard error: 1.568 on 14 degrees of freedom\nMultiple R-squared: 0.9253, Adjusted R-squared: 0.8987 \nF-statistic: 34.7 on 5 and 14 DF, p-value: 2.081e-07\n\n#calculate predictions using spline regression model\n<\/span>x_lim <- range(df$x)\nx_grid <- seq(x_lim[ 1<\/span> ], x_lim[ 2<\/span> ])\npreds <- predict(spline_fit, newdata=list(x=x_grid))\n\n#create scatter plot with spline regression predictions\n<\/span>plot(df$x, df$y, cex= 1.5<\/span> , pch= 19<\/span> )\nlines(x_grid, preds)<\/strong> <\/pre>\n![\"\"](\"https:\/\/statorials.org\/wp-content\/uploads\/2023\/08\/spline3.jpg\")
<\/p>\n Ek kaynaklar<\/strong><\/span><\/h2>\n
R’de \u00fcstel regresyon nas\u0131l ger\u00e7ekle\u015ftirilir<\/a>
R’de a\u011f\u0131rl\u0131kl\u0131 en k\u00fc\u00e7\u00fck kareler regresyonu nas\u0131l ger\u00e7ekle\u015ftirilir?<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"