{"id":1820,"date":"2019-06-13T05:12:59","date_gmt":"2019-06-12T20:12:59","guid":{"rendered":"http:\/\/141.164.34.82\/?p=1820"},"modified":"2019-06-13T06:00:55","modified_gmt":"2019-06-12T21:00:55","slug":"%ec%84%a0%ed%98%95-%eb%aa%a8%ed%98%95-2-%ec%97%b0%ec%86%8d%ed%98%95-%ec%84%a4%eb%aa%85-%eb%b3%80%ec%88%98%ec%9d%bc-%eb%95%8c-%ec%84%a0%ed%98%95%eb%b3%80%ed%99%98%ec%9d%84-%ed%86%b5%ed%95%9c-%eb%aa%a8","status":"publish","type":"post","link":"http:\/\/ds.sumeun.org\/?p=1820","title":{"rendered":"\uc120\ud615 \ubaa8\ud615 2: \uc5f0\uc18d\ud615 \uc124\uba85 \ubcc0\uc218\uc77c \ub54c \uc120\ud615\ubcc0\ud658\uc744 \ud1b5\ud55c \ubaa8\ud615 \ud589\ub82c\uc758 \uacb0\uc815"},"content":{"rendered":"

\ub4e4\uc5b4\uac00\uae30<\/h2>\n

\uc9c0\ub09c \uc2dc\uac04\uc5d0\ub294 \ud754\ud788 \ub9d0\ud558\ub294 \uc120\ud615<\/strong> \ubaa8\ud615(linear<\/strong> model)\uc758 \uc120\ud615\uc758 \uc758\ubbf8\ub97c \uc54c\uc544\ubcf4\uc558\uc2b5\ub2c8\ub2e4. \uc120\ud615<\/strong> \ubaa8\ud615\uc758 \uc120\ud615<\/strong>\uc774\ub780 \ubaa8\ud615(\\(f(x_1, x_2, \\cdots, x_p)\\))\uc774 \ub2e4\uc74c\uacfc \uac19\uc740 \ubaa8\uc218\uc758 \uc120\ud615 \ud568\uc218\uc784\uc744 \uc758\ubbf8\ud569\ub2c8\ub2e4.<\/p>\n

\\[f(x_1, x_2, \\cdots, x_p) = \\beta_0 + \\beta_1 x_1 + \\beta_2 x_2 + \\cdots \\beta_p x_p\\]<\/p>\n

\uc774\ub54c \\(x_1, x_2, \\cdots, x_p\\) \ub294 \uad00\ucc30 \ubcc0\uc218 \uadf8 \uc790\uccb4\uc77c \uc218\ub3c4 \uc788\uace0, \uad00\ucc30 \ubcc0\uc218\ub97c \ubcc0\ud658\uc2dc\ud0a8 \ubcc0\uc218(\uc608. \\(x_1^2, x_1\\cdot x_2, \\sqrt{x_1}\\) \ub4f1)\uc77c \uc218\ub3c4 \uc788\uc2b5\ub2c8\ub2e4.<\/p>\n

\uc120\ud615 \ubaa8\ud615\uc758 \uc7a5\uc810\uc740 \\(f(x_1, x_2, \\cdots, x_p) = \\vec{x}^T \\vec{\\beta}\\) \ub85c \uac04\ub2e8\ud558\uac8c \ud45c\ud604\ud560 \uc218 \uc788\uace0, \uc804\uccb4 \uc790\ub8cc\uc758 \uc608\uce21\uac12( \\(\\hat{\\vec{y}}\\) )\uc740 \\(\\textbf{X}\\vec{\\beta}\\) \ub85c \ub098\ud0c0\ub0bc \uc218 \uc788\uc2b5\ub2c8\ub2e4. <\/p>\n

\uc774\ub54c \\(\\textbf{X}\\) \ub294 \ubaa8\ub4e0 \uc0ac\ub840\uc5d0 \ub300\ud55c \\(x_1, x_2, \\cdots, x_p\\) \ub97c \ud589\ub82c\ub85c \ub9cc\ub4e0 \uac83\uc785\ub2c8\ub2e4. <\/p>\n

\\[\\textbf{X}_{ij} = x_{j}^{(i)}\\] <\/p>\n

(\ud589\ub82c\uc744 \uacb0\uc815\ud558\ub294 \uc720\uc77c\ud55c \ubc29\ubc95\uc740 \ubaa8\ub4e0 \uc6d0\uc18c\ub97c \uc815\ud574\uc8fc\ub294 \uac83\uc774\uace0, \uc704\uc758 \uc218\uc2dd\uc740 \ud589\ub82c \\(\\textbf{X}\\) \uc758 \uc784\uc758 \\(i\\) -\ud589, \\(j\\) -\uc5f4 \uc6d0\uc18c \\(\\textbf{X}_{ij}\\) \uc744 \uc54c \uc218 \uc788\ub294 \ubc29\ubc95\uc744 \uc54c\ub824 \uc900\ub2e4\uace0 \ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4. \ub9cc\uc57d \\(\\textbf{X}\\) \uc758 2\ubc88\uc9f8 \ud589, 3\ubc88\uc9f8 \uc5f4\uc758 \uc6d0\uc18c\ub97c \uc54c\uace0 \uc2f6\ub2e4\uba74 \\(i\\) \uc640 \\(j\\) \uc5d0 2\uc640 3\uc744 \ub300\uc785\ud558\uc5ec \\(x_3^{(2)}\\) \uc744 \uc5bb\uc2b5\ub2c8\ub2e4. \uadf8\ub9ac\uace0 \\(x_3^{(2)}\\) \ub294 3\ubc88\uc9f8 \ub370\uc774\ud130\uc758 2\ubc88\uc9f8 \ubcc0\uc218\ub97c \ub098\ud0c0\ub0c5\ub2c8\ub2e4.)<\/p>\n

\ubaa8\ud615 \ud589\ub82c \\(\\textbf{X}\\) \uc758 \uc815\ud558\uae30 : \uc5f0\uc18d\ud615 \ubcc0\uc218\uc758 \uacbd\uc6b0<\/h2>\n

\uad00\ucc30 \ubcc0\uc218\uc758 \ud3c9\uade0 \uc911\uc2ec\ud654(mean-centering)<\/h3>\n

\uc5ec\uae30\uc11c\ub294 \uad00\ucc30 \ubcc0\uc218\uc640 \uc120\ud615 \ubaa8\ud615\uc5d0 \ub4e4\uc5b4\uac00\ub294 \\(\\textbf{X}_{\\ \\cdot j}= x_{j}^{(\\cdot)}\\) \uac12\uc744 \uad6c\ubcc4\ud558\uae30 \uc704\ud574 \uc6b0\ub9ac\uac00 \uac00\uc9c0\uace0 \uc788\ub294 \ubcc0\uc218 \uc911 \ud558\ub098\ub97c \\(z_j\\) \ub85c \uc4f0\uaca0\uc2b5\ub2c8\ub2e4. \\(x_{j}\\) \ub294 \\(z_1\\) \uc774 \ub420 \uc218\ub3c4 \uc788\uace0, \\(\\log z_2\\) \ub610\ub294 \\(z_1 + z_2\\) \ub4f1\uc774 \ub420 \uc218 \uc788\uc2b5\ub2c8\ub2e4. <\/p>\n

\ub9cc\uc57d \uc5b4\ub5a4 \uad00\ucc30 \ubcc0\uc218 \\(z_1\\) \uac00 \uc5f0\uc18d\ud615 \ubcc0\uc218\ub77c\uace0 \ud574\ubd05\uc2dc\ub2e4. \uac00\uc7a5 \uc790\uc5f0\uc2a4\ub7ec\uc6b4 \ubc29\ubc95\uc740 \uc120\ud615 \ubaa8\ud615\uc740 \ub2e4\uc74c\uacfc \uac19\uc2b5\ub2c8\ub2e4. <\/p>\n

\\[y = \\beta_0 + \\beta_1 z_1\\]<\/p>\n

\\(x_1 = z_1\\) \uc774 \ub418\ub294 \uac83\uc774\uc8e0. \ud558\uc9c0\ub9cc \uc774 \ubc29\ubc95\uc774 \uc720\uc77c\ud55c \ubc29\ubc95\uc740 \uc544\ub2d9\ub2c8\ub2e4. \uc608\ub97c \ub4e4\uc5b4 \\(x_1=z_1 – \\bar{z_1}\\) \uc73c\ub85c \ubcc0\ud658\ud558\uc5ec \uc4f0\uae30\ub3c4 \ud569\ub2c8\ub2e4. <\/p>\n

\\[y = \\beta_0 + \\beta_1(z_1-\\bar{z_1})\\]<\/p>\n

\uc774\ub97c \uc798 \uc815\ub9ac\ud574\ubcf4\uba74 \\(y= \\beta_0 – \\beta_1\\bar{z_1} + \\beta_1 z_1\\) \uacfc \ub3d9\uc77c\ud558\uace0, \\(\\mathbb{E}[y | z_1 = \\bar{z_1}] = \\beta_0\\) \uc774 \ub418\uc5b4, \\(\\beta_0\\) \uc740 \\(z_1\\) \uc774 \ud45c\ubcf8 \ud3c9\uade0\uc77c \ub54c\uc758 \uc608\uce21\uac12\uc774 \ub418\uc5b4, \uc55e\uc758 \ubaa8\ud615 \\(y=\\beta_0+\\beta_1 z_1\\) \uc5d0\uc11c \\(\\beta_0\\) \uc758 \uc758\ubbf8(\\(z_1\\) \uc774 0\uc77c \ub54c \\(y\\) \uc758 \uae30\ub300\uac12)\uc640 \ub2ec\ub77c\uc9d1\ub2c8\ub2e4. <\/p>\n

\uc608\ub97c \ub4e4\uc5b4 \ub2e4\uc74c\uc758 \uc790\ub8cc\ub97c \ubd05\uc2dc\ub2e4.<\/p>\n

library(dplyr)\ndata(Davis, package='carData')\nDavis %>% filter(height > 100) -> Davis2\nDavis2 %>% filter(sex == "M") %>% head(10) -> Davis3\nDavis3\n<\/code><\/pre>\n
##    sex weight height repwt repht\n## 1    M     77    182    77   180\n## 2    M     68    177    70   175\n## 3    M     76    170    76   165\n## 4    M     76    167    77   165\n## 5    M     69    186    73   180\n## 6    M     71    178    71   175\n## 7    M     65    171    64   170\n## 8    M     70    175    75   174\n## 9    M     92    187   101   185\n## 10   M     76    197    75   200\n<\/pre>\n
\uc774 10\uba85\uc758 \ud559\uc0dd \uc790\ub8cc\ub97c \ud65c\uc6a9\ud558\uc5ec \ud0a4\uc640 \ubab8\ubb34\uac8c\uc758 \uad00\uacc4\ub97c \uc120\ud615 \ubaa8\ud615\uc73c\ub85c \ucd94\uc815\ud558\uace0\uc790 \ud569\ub2c8\ub2e4. \uc774\ub54c \ubab8\ubb34\uac8c\ub97c \\(y\\) \ubcc0\uc218\ub85c \ub193\uace0, \ud0a4\ub97c \\(z_1\\) \uc73c\ub85c \ub193\uc73c\uba74, \uc120\ud615 \ubaa8\ud615 \\(y = \\beta_0 + \\beta_1 x_1\\) \uc5d0\uc11c \\(x_1\\) \uc744 \\(x_1 = z_1\\) \uc73c\ub85c \ub193\uc744 \uc218\ub3c4 \uc788\uace0, \\(x_1 = z_1 – 179\\) \ub85c \ub193\uc744 \uc218\ub3c4 \uc788\uc2b5\ub2c8\ub2e4. (179\ub294 \ud559\uc0dd\ub4e4\uc758 \ud0a4 \ud3c9\uade0\uc785\ub2c8\ub2e4.)<\/p>\n
\ub450 \ubaa8\ud615 \\(y = \\beta_0 + \\beta_1 z_1\\) \uacfc \\(y = \\beta_0 + \\beta_1 (z_1 – 179)\\) \ub294 \ubaa8\ub450 \ubaa8\uc218\uc758 \uc120\ud615 \ubaa8\ud615\uc774\uace0, \ub3d9\uc77c\ud55c \ubaa8\ud615\uc744 \uc870\uae08 \ub2e4\ub974\uac8c \ub098\ud0c0\ub0b8 \uac83 \ubfd0\uc774\uc9c0\ub9cc, \\(\\beta_0\\) \uc758 \uc758\ubbf8\ub294 \ub2ec\ub77c\uc9d1\ub2c8\ub2e4.<\/p>\n
fit0 <- lm(weight ~ height, data=Davis3)\nfitCentered <- lm(weight ~ I(height - mean(height)), data=Davis3)\nsummary(fit0)\n<\/code><\/pre>\n
## \n## Call:\n## lm(formula = weight ~ height, data = Davis3)\n## \n## Residuals:\n##    Min     1Q Median     3Q    Max \n## -6.991 -4.853 -2.789  3.956 15.725 \n## \n## Coefficients:\n##             Estimate Std. Error t value Pr(>|t|)\n## (Intercept)  23.0939    48.9303   0.472    0.650\n## height        0.2844     0.2730   1.042    0.328\n## \n## Residual standard error: 7.507 on 8 degrees of freedom\n## Multiple R-squared:  0.1194, Adjusted R-squared:  0.00935 \n## F-statistic: 1.085 on 1 and 8 DF,  p-value: 0.328\n<\/pre>\n
summary(fitCentered)\n<\/code><\/pre>\n
## \n## Call:\n## lm(formula = weight ~ I(height - mean(height)), data = Davis3)\n## \n## Residuals:\n##    Min     1Q Median     3Q    Max \n## -6.991 -4.853 -2.789  3.956 15.725 \n## \n## Coefficients:\n##                          Estimate Std. Error t value Pr(>|t|)    \n## (Intercept)               74.0000     2.3740  31.172 1.22e-09 ***\n## I(height - mean(height))   0.2844     0.2730   1.042    0.328    \n## ---\n## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1\n## \n## Residual standard error: 7.507 on 8 degrees of freedom\n## Multiple R-squared:  0.1194, Adjusted R-squared:  0.00935 \n## F-statistic: 1.085 on 1 and 8 DF,  p-value: 0.328\n<\/pre>\n
\uc704\uc758 \uacb0\uacfc\ub97c \ubcf4\uba74 \uc124\uba85\ubcc0\uc218\ub97c \ud3c9\uade0\uc911\uc2ec\ud654\ud574\ub3c4 \\(\\beta_1\\) \uc758 \ucd94\uc815\uac12\uc774\ub098 \uc2e0\ub8b0\uad6c\uac04\uc740 \ubcc0\ud558\uc9c0 \uc54a\uc74c\uc744 \ud655\uc778\ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4. \ud558\uc9c0\ub9cc \\(\\beta_0\\) \uc758 \ucd94\uc815\uac12\uacfc \uc2e0\ub8b0\uad6c\uac04\uc740 \ub2ec\ub77c\uc9d1\ub2c8\ub2e4. \\(\\beta_0\\) \uc758 \uc758\ubbf8\ub294 \ud0a4\uac00 179\uc77c \ub54c \ucd94\uc815\ub418\ub294 \uccb4\uc911\uc758 \ud3c9\uade0\uc785\ub2c8\ub2e4. (\uc5ec\uae30\uc11c 179\uc740 \ud3c9\uade0\uc758 \ucd94\uc815\uac12\uc774\uc9c0 \uc815\ud655\ud558\uac8c \ud0a4\uc758 \ud3c9\uade0\uc774\ub77c\uace0 \ud560 \uc218 \uc5c6\uae30 \ub54c\ubb38\uc5d0 \ud3c9\uade0 \ud0a4\uc5d0\uc11c \ucd94\uc815\ub418\ub294 \uccb4\uc911\uc758 \ud3c9\uade0\uc774\ub77c\uace0 \ub9d0\ud560 \uc21c \uc5c6\uc2b5\ub2c8\ub2e4. \ubb3c\ub860 \ud45c\ubcf8\uc758 \ud06c\uae30\uac00 \ucee4\uc9c8 \uc218\ub85d \ud45c\ubcf8\ud3c9\uade0\uacfc \ubaa8\ud3c9\uade0\uc758 \ucc28\uc774\uac00 \uc791\uc544\uc9d1\ub2c8\ub2e4.)<\/p>\n
confint(fit0)\n<\/code><\/pre>\n
##                   2.5 %      97.5 %\n## (Intercept) -89.7395203 135.9273510\n## height       -0.3452205   0.9140036\n<\/pre>\n
confint(fitCentered)\n<\/code><\/pre>\n
##                               2.5 %     97.5 %\n## (Intercept)              68.5256328 79.4743672\n## I(height - mean(height)) -0.3452205  0.9140036\n<\/pre>\n
\uc544\ub798 \uadf8\ub798\ud504\uc5d0\uc11c \uc790\ub8cc\uc758 \uc704\uce58\uc640 \ud68c\uadc0\uc120, \uadf8\ub9ac\uace0 \\(\\textrm{height}\\) \uc5d0 \ub530\ub978 \\(\\textrm{weight}\\) \uc758 \ucd94\uc815\uac12\uc744 \ud655\uc778\ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4. \ud3c9\uade0 \uc911\uc2ec\ud654\ub97c \ud558\uc9c0 \uc54a\uc740 \uacbd\uc6b0\uc5d0 \\(\\beta_0\\) \uc758 \uc2e0\ub8b0\uad6c\uac04\uc774 \uc5c4\uccad\ub098\uac8c \ucee4\uc9c0\ub294 \uc774\uc720\ub97c \ud655\uc778\ud560 \uc218 \uc788\uc744 \uac83\uc785\ub2c8\ub2e4. <\/p>\n
library(ggplot2)\ndat = data.frame(height=seq(0,200 , 1))\ndatPred = as.data.frame(predict(fit0, newdata=dat, interval='prediction'))\ndatPred$height = dat$height\nggplot(Davis3, aes(x=height, y=weight)) + \n  geom_ribbon(data = datPred, aes(y=fit, ymin=lwr, ymax=upr), alpha=0.2) + \n  geom_line(data=datPred, aes(x=height, y=fit))+ \n  geom_point() \n<\/code><\/pre>\n
## Warning: Ignoring unknown aesthetics: y\n<\/pre>\n
\"plot<\/p>\n
\uc124\uba85\ubcc0\uc218\uc758 \ud45c\uc900\ud654<\/h3>\n
\uc124\uba85\ubcc0\uc218\uc758 \ud3c9\uade0\uc911\uc2ec\ud654\uc5d0 \ud55c \ub2e8\uacc4\ub97c \ub354\ud574\uc11c \ud45c\uc900\ud654\ub97c \uc2dc\ud0ac \uc218\ub3c4 \uc788\uc2b5\ub2c8\ub2e4. \ud45c\uc900\ud654\ub294 \ub2e4\uc74c\uacfc \uac19\uc774 \uc4f8 \uc218 \uc788\uc2b5\ub2c8\ub2e4. <\/p>\n
\\[x_1 = \\frac{z_1 – \\bar{z_1}}{\\textrm{sd}(z_1)}\\]<\/p>\n
\uc5ec\uae30\uc11c \\(\\bar{z_1}\\) \uacfc \\(\\textrm{sd}(z_1)\\) \uc740 \\(z_1\\) \uc758 \ud45c\ubcf8 \ud3c9\uade0\uacfc \ud45c\ubcf8 \ud45c\uc900\ud3b8\ucc28\uc785\ub2c8\ub2e4. <\/p>\n
\uc774\ub807\uac8c \\(x_1\\) \uc744 \uc124\uc815\ud558\uba74 \\(y = \\beta_0 + \\beta_1 x_1\\) \uc5d0\uc11c \\(\\beta_1\\) \uc758 \uc758\ubbf8\uac00 \\(z_1\\) \uc774 1\ud45c\uc900\ud3b8\ucc28\ub9cc\ud07c \uc99d\uac00\ud560 \ub54c, \uae30\ub300\ub418\ub294 \\(y\\) \uac12\uc758 \uc99d\uac00\ub85c \ud574\uc11d\ud560 \uc218 \uc788\ub2e4\ub294 \uc7a5\uc810\uc774 \uc788\uc2b5\ub2c8\ub2e4. \\(z_1\\) \uc758 \uac12\uc758 \uc758\ubbf8\ub97c \ud574\uc11d\ud558\uae30 \uc5b4\ub824\uc6b4 \uacbd\uc6b0\uc5d0 \uc720\uc6a9\ud55c \ubc29\ubc95\uc785\ub2c8\ub2e4. \uc608\ub97c \ub4e4\uc5b4 \ubd05\uc2dc\ub2e4. \\(z_1\\) \uc774 \ud0a4\ub97c \ub098\ud0c0\ub0b4\uace0 \ubbf8\ud130 \ub2e8\uc704\ub85c \uce21\uc815\ub418\uc5c8\ub2e4\uba74, \\(x_1=z_1\\) \uc73c\ub85c \ub193\uc544\ub3c4 \\(\\beta_1\\) \uc758 \uc758\ubbf8\uac00 \ud0a4\uac00 \\(1\\textrm{m}\\) \uc99d\uac00\ud560 \ub54c \\(y\\) \uc758 \uae30\ub300 \uc99d\uac00\ub7c9\uc73c\ub85c \uc27d\uac8c \uc774\ud574\ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4. \ud558\uc9c0\ub9cc \\(z_1\\) \uc774 \uc5b4\ub5a4 \uc124\ubb38\uc870\uc0ac\uc758 \ub9ac\ucee4\ud2b8 \ucc99\ub3c4 \uc751\ub2f5\uc774\ub77c\uace0 \ud55c\ub2e4\uba74, \uc774\ub54c 1\uc758 \uc99d\uac00\uac00 \uc815\ud655\ud788 \uc5b4\ub5a4 \uc758\ubbf8\uac00 \uc788\ub294\uc9c0 \ud30c\uc545\ud558\uae30\uac00 \uc27d\uc9c0\ub294 \uc54a\uc2b5\ub2c8\ub2e4. \uc774\ub7f4 \ub54c \\(x_1 = \\frac{z_1 – \\bar{z_1}}{\\textrm{sd}(z_1)}\\) \uc73c\ub85c \ub193\uc73c\uba74, \uc751\ub2f5 \uc9d1\ub2e8 \ub0b4\uc5d0\uc11c \uc5bc\ub9c8\ub098 \ud070 \ucc28\uc774\uc778\uc9c0\ub97c \ub300\ub7b5\uc801\uc73c\ub85c \uc774\ud574\ud560 \uc218 \uc788\uac8c \ub429\ub2c8\ub2e4. (\uc774\ub54c \\(\\textrm{sd}(z_1)\\) \uc740 \ud45c\ubcf8\ud45c\uc900\ud3b8\ucc28\ub85c \uc815\ud655\ud55c \ubaa8\ud45c\uc900\ud3b8\ucc28\uac00 \uc544\ub2c8\uae30 \ub54c\ubb38\uc5d0 \uc77c\ubc18\uc801\uc778 \uacc4\uc218 \ucd94\uc815\ub7c9\uc5d0 \ube44\ud574 \ubd88\ud655\uc2e4\uc131\uc774 \ucee4\uc9c0\uc9c0\ub9cc \ub300\ubd80\ubd84\uc758 \uacbd\uc6b0 \ud06c\uac8c \uc2e0\uacbd\uc744 \uc368\uc57c \ud560 \uc815\ub3c4\ub294 \uc544\ub2d0 \uac83\uc785\ub2c8\ub2e4.)<\/p>\n
fit0 <- lm(weight ~ height, data=Davis3)\nfitStd <- lm(weight ~ I((height - mean(height))\/sd(height)), data=Davis3)\nsummary(fit0)\n<\/code><\/pre>\n
## \n## Call:\n## lm(formula = weight ~ height, data = Davis3)\n## \n## Residuals:\n##    Min     1Q Median     3Q    Max \n## -6.991 -4.853 -2.789  3.956 15.725 \n## \n## Coefficients:\n##             Estimate Std. Error t value Pr(>|t|)\n## (Intercept)  23.0939    48.9303   0.472    0.650\n## height        0.2844     0.2730   1.042    0.328\n## \n## Residual standard error: 7.507 on 8 degrees of freedom\n## Multiple R-squared:  0.1194, Adjusted R-squared:  0.00935 \n## F-statistic: 1.085 on 1 and 8 DF,  p-value: 0.328\n<\/pre>\n
summary(fitStd)\n<\/code><\/pre>\n
## \n## Call:\n## lm(formula = weight ~ I((height - mean(height))\/sd(height)), \n##     data = Davis3)\n## \n## Residuals:\n##    Min     1Q Median     3Q    Max \n## -6.991 -4.853 -2.789  3.956 15.725 \n## \n## Coefficients:\n##                                       Estimate Std. Error t value Pr(>|t|)\n## (Intercept)                             74.000      2.374  31.172 1.22e-09\n## I((height - mean(height))\/sd(height))    2.606      2.502   1.042    0.328\n##                                          \n## (Intercept)                           ***\n## I((height - mean(height))\/sd(height))    \n## ---\n## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1\n## \n## Residual standard error: 7.507 on 8 degrees of freedom\n## Multiple R-squared:  0.1194, Adjusted R-squared:  0.00935 \n## F-statistic: 1.085 on 1 and 8 DF,  p-value: 0.328\n<\/pre>\n
\uc704\uc758 \uacb0\uacfc\ub97c \ubcf4\uba74 \ud0a4\uac00 1\ud45c\uc900\ud3b8\ucc28\ub9cc\ud07c \uc99d\uac00\ud560 \ub54c \uae30\ub300\ub418\ub294 \uccb4\uc911 \uc99d\uac00\ub7c9\uc740 2.606 kg\uc784\uc744 \ud655\uc778\ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4. <\/p>\n
\uc124\uba85\ubcc0\uc218\uc640 \uacb0\uacfc\ubcc0\uc218\uc758 \ud45c\uc900\ud654<\/h3>\n
\ud754\ud788 \ud45c\uc900\ud654\ud68c\uadc0\uacc4\uc218(standardized regression coefficient)<\/strong> \ub610\ub294 \ubca0\ud0c0 \uacc4\uc218(beta coefficient)<\/strong>\ub77c\uace0 \ubd88\ub9ac\ub294 \ud68c\uadc0 \uacc4\uc218\ub294 \uc124\uba85\ubcc0\uc218\uc640 \uacb0\uacfc\ubcc0\uc218\ub97c \ubaa8\ub450 \ud45c\uc900\ud654 \uc2dc\ucf30\uc744 \ub54c \uc5bb\ub294 \uacc4\uc218\uc785\ub2c8\ub2e4.<\/p>\n
\\[x_1 = \\frac{z_1 – \\bar{z_1}}{\\textrm{sd}(z_1)}, \\ \\ \\tilde{y} = \\frac{y – \\bar{y}}{\\textrm{sd}(y)}\\]<\/p>\n
\uc774\ub807\uac8c \uc124\uba85\ubcc0\uc218\uc640 \uacb0\uacfc\ubcc0\uc218\ub97c \ubcc0\ud658\ud55c \ud6c4 \ud68c\uadc0\uc2dd \\(\\tilde{y} = \\beta_0 + \\beta_1 x_1\\) \uc73c\ub85c \ud68c\uadc0\ubd84\uc11d\uc744 \ud588\uc744 \ub54c, \\(\\beta_1\\) \uc758 \uc758\ubbf8\ub294 \\(z_1\\) \uc744 \ud45c\uc900\ud3b8\ucc28 1\ub9cc\ud07c \uc99d\uac00\uc2dc\ucf30\uc744 \ub54c \uae30\ub300\ub418\ub294 \uacb0\uacfc \ubcc0\uc218 \\(y\\) \uc758 \uc99d\uac00\ub7c9\uc744 \\(y\\) \uc758 \ud45c\uc900\ud3b8\ucc28\ub85c \ub098\ub208 \uac12\uc785\ub2c8\ub2e4. \uc608\ub97c \ub4e4\uc5b4 \\(\\beta_1=1.5\\) \ub77c\uba74 \\(z_1\\) \uac00 1\ud45c\uc900\ud3b8\ucc28\ub9cc\ud07c \uc99d\uac00\ud560 \ub54c, \\(y\\) \ub294 1.5 \ud45c\uc900\ud3b8\ucc28\ub9cc\ud07c \uc99d\uac00\ud558\ub294 \uac83\uc785\ub2c8\ub2e4.<\/p>\n
\ud45c\uc900\ud654\ud68c\uadc0\uacc4\uc218\ub294 \\(z_1\\) \uc640 \\(y\\) \uac00 \ubaa8\ub450 \uc815\ud655\ud55c \uc758\ubbf8\ub97c \uc54c \uc218 \uc5c6\uc744 \ub54c \uc720\uc6a9\ud558\uac8c \uc0ac\uc6a9\ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4.<\/p>\n
fitbStd <- lm(I((weight-mean(weight))\/sd(weight)) ~ I((height - mean(height))\/sd(height)), data=Davis3)\nsummary(fitbStd)\n<\/code><\/pre>\n
## \n## Call:\n## lm(formula = I((weight - mean(weight))\/sd(weight)) ~ I((height - \n##     mean(height))\/sd(height)), data = Davis3)\n## \n## Residuals:\n##     Min      1Q  Median      3Q     Max \n## -0.9268 -0.6434 -0.3698  0.5245  2.0848 \n## \n## Coefficients:\n##                                        Estimate Std. Error t value\n## (Intercept)                           5.266e-17  3.147e-01   0.000\n## I((height - mean(height))\/sd(height)) 3.456e-01  3.318e-01   1.042\n##                                       Pr(>|t|)\n## (Intercept)                              1.000\n## I((height - mean(height))\/sd(height))    0.328\n## \n## Residual standard error: 0.9953 on 8 degrees of freedom\n## Multiple R-squared:  0.1194, Adjusted R-squared:  0.00935 \n## F-statistic: 1.085 on 1 and 8 DF,  p-value: 0.328\n<\/pre>\n
R\uc5d0\ub294 \ud45c\uc900\ud654\ub97c \ud568\uc218\ub294 scale()<\/code>\uc785\ub2c8\ub2e4. \ub2e4\uc74c\uc758 \uacb0\uacfc\ub294 \uc704\uc758 \uacb0\uacfc\uc640 \ub3d9\uc77c\ud568\uc744 \uc54c \uc218 \uc788\uc2b5\ub2c8\ub2e4.<\/p>\n
summary(lm(scale(weight) ~ scale(height), data=Davis3))\n<\/code><\/pre>\n
## \n## Call:\n## lm(formula = scale(weight) ~ scale(height), data = Davis3)\n## \n## Residuals:\n##     Min      1Q  Median      3Q     Max \n## -0.9268 -0.6434 -0.3698  0.5245  2.0848 \n## \n## Coefficients:\n##                Estimate Std. Error t value Pr(>|t|)\n## (Intercept)   5.266e-17  3.147e-01   0.000    1.000\n## scale(height) 3.456e-01  3.318e-01   1.042    0.328\n## \n## Residual standard error: 0.9953 on 8 degrees of freedom\n## Multiple R-squared:  0.1194, Adjusted R-squared:  0.00935 \n## F-statistic: 1.085 on 1 and 8 DF,  p-value: 0.328\n<\/pre>\n
\uc815\ub9ac<\/h3>\n
\uc5ec\uae30\uc11c\ub294 \uc5f0\uc18d\ud615 \uad00\ucc30\ubcc0\uc218\ub85c \uacb0\uacfc\ubcc0\uc218\ub97c \uc608\uce21(\uc124\uba85)\ud558\ub294 \uc120\ud615 \ubaa8\ud615\uc744 \ub9cc\ub4e4 \ub54c \ubaa8\ud615 \ud589\ub82c\uc744 \uacb0\uc815\ud558\ub294 \uc5ec\ub7ec \uac00\uc9c0 \ub2e4\ub978 \ubc29\ubc95\uc5d0 \ub300\ud574 \uc54c\uc544\ubcf4\uc558\uc2b5\ub2c8\ub2e4. \uad6c\uccb4\uc801\uc73c\ub85c \ubcc0\uc218\uc758 \ud3c9\uade0 \uc911\uc2ec\ud654<\/strong>, \ud45c\uc900\ud654<\/strong>, \uadf8\ub9ac\uace0 \uc124\uba85\ubcc0\uc218\uc640 \uacb0\uacfc\ubcc0\uc218 \ubaa8\ub450\uc758 \ud45c\uc900\ud654<\/strong>\ub97c \uc0ac\uc6a9\ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4. \uc774\ub4e4\uc740 \ubaa8\ub450 \uc124\uba85\ubcc0\uc218\uc758 \uc120\ud615 \ud568\uc218<\/strong>\uc774\ub77c\ub294 \uacf5\ud1b5\uc810\uc774 \uc788\uc2b5\ub2c8\ub2e4. <\/p>\n","protected":false},"excerpt":{"rendered":"
\ub4e4\uc5b4\uac00\uae30 \uc9c0\ub09c \uc2dc\uac04\uc5d0\ub294 \ud754\ud788 \ub9d0\ud558\ub294 \uc120\ud615 \ubaa8\ud615(linear model)\uc758 \uc120\ud615\uc758 \uc758\ubbf8\ub97c \uc54c\uc544\ubcf4\uc558\uc2b5\ub2c8\ub2e4. \uc120\ud615 \ubaa8\ud615\uc758 \uc120\ud615\uc774\ub780 \ubaa8\ud615(\\(f(x_1, x_2, \\cdots, x_p)\\))\uc774 \ub2e4\uc74c\uacfc \uac19\uc740 \ubaa8\uc218\uc758 \uc120\ud615 \ud568\uc218\uc784\uc744 \uc758\ubbf8\ud569\ub2c8\ub2e4. \\[f(x_1, x_2, \\cdots, x_p) = \\beta_0 + \\beta_1 x_1 + \\beta_2 x_2 + \\cdots \\beta_p x_p\\] \uc774\ub54c \\(x_1, x_2, \\cdots, x_p\\) \ub294 \uad00\ucc30 \ubcc0\uc218 \uadf8 \uc790\uccb4\uc77c \uc218\ub3c4 \uc788\uace0, \uad00\ucc30 \ubcc0\uc218\ub97c \ubcc0\ud658\uc2dc\ud0a8 […]<\/p>\n","protected":false},"author":1,"featured_media":1823,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":[],"categories":[28,319,320],"tags":[314,324,327,329,321,328,322,325,326,331,332,330],"jetpack_featured_media_url":"http:\/\/ds.sumeun.org\/wp-content\/uploads\/2019\/06\/lm_02.png","_links":{"self":[{"href":"http:\/\/ds.sumeun.org\/index.php?rest_route=\/wp\/v2\/posts\/1820"}],"collection":[{"href":"http:\/\/ds.sumeun.org\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"http:\/\/ds.sumeun.org\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"http:\/\/ds.sumeun.org\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"http:\/\/ds.sumeun.org\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=1820"}],"version-history":[{"count":9,"href":"http:\/\/ds.sumeun.org\/index.php?rest_route=\/wp\/v2\/posts\/1820\/revisions"}],"predecessor-version":[{"id":1830,"href":"http:\/\/ds.sumeun.org\/index.php?rest_route=\/wp\/v2\/posts\/1820\/revisions\/1830"}],"wp:featuredmedia":[{"embeddable":true,"href":"http:\/\/ds.sumeun.org\/index.php?rest_route=\/wp\/v2\/media\/1823"}],"wp:attachment":[{"href":"http:\/\/ds.sumeun.org\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=1820"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"http:\/\/ds.sumeun.org\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=1820"},{"taxonomy":"post_tag","embeddable":true,"href":"http:\/\/ds.sumeun.org\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=1820"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}