<\/p>\n
White testi,<\/strong> bir regresyon modelinde de\u011fi\u015fen varyans\u0131n<\/a> mevcut olup olmad\u0131\u011f\u0131n\u0131 belirlemek i\u00e7in kullan\u0131l\u0131r.<\/span><\/p>\n De\u011fi\u015fen varyans, bir yan\u0131t de\u011fi\u015fkeninin<\/a> farkl\u0131 seviyelerindeki art\u0131klar\u0131n<\/a> e\u015fit olmayan da\u011f\u0131l\u0131m\u0131n\u0131 ifade eder; bu, art\u0131klar\u0131n yan\u0131t de\u011fi\u015fkeninin her d\u00fczeyinde e\u015fit \u015fekilde da\u011f\u0131ld\u0131\u011f\u0131varsay\u0131m\u0131n\u0131<\/a> ihlal eder.<\/span><\/p>\n A\u015fa\u011f\u0131daki ad\u0131m ad\u0131m \u00f6rnek, de\u011fi\u015fen varyans\u0131n belirli bir regresyon modelinde bir sorun olup olmad\u0131\u011f\u0131n\u0131 belirlemek i\u00e7in Python’da White testinin nas\u0131l ger\u00e7ekle\u015ftirilece\u011fini g\u00f6sterir.<\/span><\/p>\n Bu \u00f6rnekte mtcars<\/strong> veri k\u00fcmesini kullanarak \u00e7oklu do\u011frusal regresyon modelini<\/a> uygulayaca\u011f\u0131z.<\/span><\/p>\n A\u015fa\u011f\u0131daki kod, bu veri k\u00fcmesinin bir pandas DataFrame’e nas\u0131l y\u00fcklenece\u011fini g\u00f6sterir:<\/span><\/p>\n Daha sonra, yan\u0131t de\u011fi\u015fkeni olarak mpg’yi<\/strong> ve iki \u00f6ng\u00f6r\u00fcc\u00fc de\u011fi\u015fken olarak disp<\/strong> ve hp’yi<\/strong> kullanarak bir regresyon modeli uygulayaca\u011f\u0131z:<\/span><\/p>\n Daha sonra, regresyon modelinde de\u011fi\u015fen varyans\u0131n mevcut olup olmad\u0131\u011f\u0131n\u0131 belirlemek amac\u0131yla White testini ger\u00e7ekle\u015ftirmek i\u00e7in statsmodels paketindeki het_white()<\/a> fonksiyonunu kullanaca\u011f\u0131z:<\/span><\/p>\n Sonucun nas\u0131l yorumlanaca\u011f\u0131 a\u015fa\u011f\u0131da a\u00e7\u0131klanm\u0131\u015ft\u0131r:<\/span><\/p>\n White testi a\u015fa\u011f\u0131daki bo\u015f ve alternatif hipotezleri kullan\u0131r:<\/span><\/p>\n P de\u011feri 0,05’ten k\u00fc\u00e7\u00fck olmad\u0131\u011f\u0131ndan s\u0131f\u0131r hipotezini reddedemiyoruz.<\/span><\/p>\n Bu, regresyon modelinde de\u011fi\u015fen varyans\u0131n mevcut oldu\u011funu iddia etmek i\u00e7in yeterli kan\u0131ta sahip olmad\u0131\u011f\u0131m\u0131z anlam\u0131na gelir.<\/span><\/p>\n White testinin s\u0131f\u0131r hipotezini reddetmeyi ba\u015faramazsan\u0131z, o zaman heteroskedastisite mevcut de\u011fildir ve orijinal regresyonun sonucunu yorumlamaya devam edebilirsiniz.<\/span><\/p>\n Ancak s\u0131f\u0131r hipotezini reddederseniz bu, de\u011fi\u015fen varyans oldu\u011fu anlam\u0131na gelir. Bu durumda regresyon \u00e7\u0131kt\u0131 tablosunda g\u00f6r\u00fcnt\u00fclenen standart hatalar g\u00fcvenilir olmayabilir.<\/span><\/p>\n Bu sorunu \u00e7\u00f6zmenin iki yayg\u0131n yolu vard\u0131r:<\/span><\/p>\n 1. Yan\u0131t de\u011fi\u015fkenini d\u00f6n\u00fc\u015ft\u00fcr\u00fcn.<\/strong><\/span><\/p>\n Yan\u0131t de\u011fi\u015fkeni \u00fczerinde bir d\u00f6n\u00fc\u015f\u00fcm ger\u00e7ekle\u015ftirmeyi deneyebilirsiniz; \u00f6rne\u011fin yan\u0131t de\u011fi\u015fkeninin g\u00fcnl\u00fc\u011f\u00fcn\u00fc, karek\u00f6k\u00fcn\u00fc veya k\u00fcp k\u00f6k\u00fcn\u00fc<\/a> alabilirsiniz. Bu genellikle heteroskedastisitenin ortadan kalkmas\u0131yla sonu\u00e7lan\u0131r.<\/span><\/p>\n 2. A\u011f\u0131rl\u0131kl\u0131 regresyon kullan\u0131n.<\/strong><\/span><\/p>\n A\u011f\u0131rl\u0131kl\u0131 regresyon, her veri noktas\u0131na, uydurulan de\u011ferin varyans\u0131na dayal\u0131 olarak bir a\u011f\u0131rl\u0131k atar. Temel olarak bu, daha y\u00fcksek varyansa sahip veri noktalar\u0131na d\u00fc\u015f\u00fck a\u011f\u0131rl\u0131k vererek bunlar\u0131n kalan karelerini azalt\u0131r. Uygun a\u011f\u0131rl\u0131klar kullan\u0131ld\u0131\u011f\u0131nda de\u011fi\u015fen varyans sorunu ortadan kald\u0131r\u0131labilir.<\/span><\/p>\n A\u015fa\u011f\u0131daki e\u011fitimler Python’da do\u011frusal regresyon hakk\u0131nda ek bilgi sa\u011flar:<\/span><\/p>\n Python’da Do\u011frusal Regresyon \u0130\u00e7in Tam Bir K\u0131lavuz<\/a> White testi, bir regresyon modelinde de\u011fi\u015fen varyans\u0131n mevcut olup olmad\u0131\u011f\u0131n\u0131 belirlemek i\u00e7in kullan\u0131l\u0131r. De\u011fi\u015fen varyans, bir yan\u0131t de\u011fi\u015fkeninin farkl\u0131 seviyelerindeki art\u0131klar\u0131n e\u015fit olmayan da\u011f\u0131l\u0131m\u0131n\u0131 ifade eder; bu, art\u0131klar\u0131n yan\u0131t de\u011fi\u015fkeninin her d\u00fczeyinde e\u015fit \u015fekilde da\u011f\u0131ld\u0131\u011f\u0131varsay\u0131m\u0131n\u0131 ihlal eder. A\u015fa\u011f\u0131daki ad\u0131m ad\u0131m \u00f6rnek, de\u011fi\u015fen varyans\u0131n belirli bir regresyon modelinde bir sorun olup olmad\u0131\u011f\u0131n\u0131 belirlemek i\u00e7in Python’da White […]<\/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-2470","post","type-post","status-publish","format-standard","hentry","category-rehber"],"yoast_head":"\n 1. Ad\u0131m: Verileri y\u00fckleyin<\/strong><\/h3>\n
from<\/span> sklearn. linear_model<\/span> import<\/span> LinearRegression\nfrom<\/span> statsmodels. stats<\/span> . diagnostic<\/span> import<\/span> het_white\nimport<\/span> statsmodels. api<\/span> as<\/span> sm\nimport<\/span> pandas as<\/span> pd\n\n#define URL where dataset is located\n<\/span>url = \"https:\/\/raw.githubusercontent.com\/Statorials\/Python-Guides\/main\/mtcars.csv\"\n\n#read in data\n<\/span>data = pd. read_csv<\/span> (url)\n\n#view summary of data\n<\/span>data. info<\/span> ()\n\n<class 'pandas.core.frame.DataFrame'>\nRangeIndex: 32 entries, 0 to 31\nData columns (total 12 columns):\n # Column Non-Null Count Dtype \n--- ------ -------------- ----- \n 0 model 32 non-null object \n 1 mpg 32 non-null float64\n 2 cyl 32 non-null int64 \n 3 disp 32 non-null float64\n 4 hp 32 non-null int64 \n 5 drat 32 non-null float64\n 6 wt 32 non-null float64\n 7 qsec 32 non-null float64\n 8 vs 32 non-null int64 \n 9 am 32 non-null int64 \n 10 gear 32 non-null int64 \n 11 carb 32 non-null int64 \ndtypes: float64(5), int64(6), object(1)\n<\/strong><\/pre>\n Ad\u0131m 2: Regresyon modelini yerle\u015ftirin<\/strong><\/h3>\n
#define response variable\n<\/span>y = data[' mpg<\/span> ']\n\n#define predictor variables\n<\/span>x = data[[' disp<\/span> ', ' hp<\/span> ']]\n\n#add constant to predictor variables\n<\/span>x = sm. add_constant<\/span> (x)\n\n#fit regression model\n<\/span>model = sm. OLS<\/span> (y,x). fit<\/span> ()<\/strong><\/pre>\n Ad\u0131m 3: White’\u0131n testini ger\u00e7ekle\u015ftirin<\/strong><\/span><\/h3>\n
#perform White's test\n<\/span>white_test = het_white(model. resid<\/span> , model. model<\/span> . exog<\/span> )\n\n#define labels to use for output of White's test\n<\/span>labels = ['Test Statistic', 'Test Statistic p-value', 'F-Statistic', 'F-Test p-value']\n\n#print results of White's test\nprint<\/span><\/span> (dict(zip(labels, white_test)))\n\n{'Test Statistic': 7.076620330416624, 'Test Statistic p-value': 0.21500404394263936,\n 'F-Statistic': 1.4764621093131864, 'F-Test p-value': 0.23147065943879694}<\/strong><\/pre>\n\n
\n
Sonra ne yapaca\u011f\u0131z<\/strong><\/h3>\n
Ek kaynaklar<\/strong><\/span><\/h3>\n
Python’da Art\u0131k Arsa Nas\u0131l Olu\u015fturulur<\/a>
Python’da VIF nas\u0131l hesaplan\u0131r<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"