{"id":656,"date":"2020-12-05T14:42:01","date_gmt":"2020-12-05T05:42:01","guid":{"rendered":"http:\/\/cedartrees.co.kr\/?p=656"},"modified":"2021-01-09T19:18:43","modified_gmt":"2021-01-09T10:18:43","slug":"sklearn-classification","status":"publish","type":"post","link":"http:\/\/blog.cedartrees.co.kr\/index.php\/2020\/12\/05\/sklearn-classification\/","title":{"rendered":"\ub85c\uc9c0\uc2a4\ud2f1 \ud68c\uadc0(Logistic Regression) \ub2e4\uc911\ubd84\ub958"},"content":{"rendered":"\n

\uc774\uc804\uae4c\uc9c0\ub294 age, weight, oxygen, runtime, runpulse, rstpulse \ub370\uc774\ud130\ub97c \ud1b5\ud574\uc11c maxpulse\ub97c \uc608\uce21\ud558\ub294 \uc608\uc81c\ub97c \uc0b4\ud3b4\ubd24\uc2b5\ub2c8\ub2e4. maxpulse\ub294 \uc5f0\uc18d\ub41c \uac12\uc744 \uac00\uc9c0\uace0 \uc788\ub294 \uc218\uce58-\uc5f0\uc18d\ud615 \ub370\uc774\ud130\uc785\ub2c8\ub2e4. <\/p>\n\n\n\n

dataset = pd.read_csv('.\/fitness.csv')\ndataset.head()<\/pre>\n\n\n\n
\"\"<\/figure>\n\n\n\n
\ub370\uc774\ud130\uc758 \ud2b9\uc9d5\uc740 \uc544\ub798\uc640 \uac19\uc2b5\ub2c8\ub2e4.<\/p>\n\n\n\n
dataset.info()<\/pre>\n\n\n\n
RangeIndex: 31 entries, 0 to 30
Data columns (total 7 columns):
# Column Non-Null Count Dtype
— —— ————– —–
0 age 31 non-null int64
1 weight 31 non-null float64
2 oxygen 31 non-null float64
3 runtime 31 non-null float64
4 runpulse 31 non-null int64
5 rstpulse 31 non-null int64
6 maxpulse 31 non-null int64
dtypes: float64(3), int64(4)
memory usage: 1.8 KB<\/p>\n\n\n\n
\uc55e\uc73c\ub85c \ud558\ub824\uace0 \ud558\ub294 \ubd80\ubd84\uc740 weight, oxygen, runtime, runpulse, rstpulse, maxpulse \uce7c\ub7fc \uc815\ubcf4\ub97c \ud1b5\ud574\uc11c \uc774 \uc0ac\ub78c\uc774 \uc5b4\ub290 \ub098\uc774\ub300\uc778\uac00\ub97c \uc608\uce21\ud558\ub294 \ubd84\ub958\uc758 \ubb38\uc81c\ub97c \ud574\uacb0\ud574\ubcf4\ub3c4\ub85d \ud558\uaca0\uc2b5\ub2c8\ub2e4. <\/p>\n\n\n\n
\uadf8\ub7ec\ub098 \uc544\uc27d\uac8c\ub3c4 \uc608\uc81c \ub370\uc774\ud130\ub294 \ubc94\uc8fc\ud615 \uceec\ub7fc\uc740 \uc5c6\uace0 \ubaa8\ub450 \uc218\uce58\ud615-\uc5f0\uc18d\ud615 \ub370\uc774\ud130\ub4e4\uc785\ub2c8\ub2e4. \uadf8\ub807\uae30 \ub54c\ubb38\uc5d0 \ubc94\uc8fc\ud615-\uc21c\uc11c\ud615 \ub370\uc774\ud130 \uceec\ub7fc\uc744 \ud558\ub098 \ub9cc\ub4e4\uc5b4\uc918\uc57c\ud569\ub2c8\ub2e4. \uc544\ub798\uc640 \uac19\uc740 \ubc94\uc8fc\ud615 \ub370\uc774\ud130 \uceec\ub7fc\uc744 \ud558\ub098 \ub9cc\ub4e4\uaca0\uc2b5\ub2c8\ub2e4. 
\ub098\uc774\uac00 40\ub300 \uc774\ud558\uc774\uba74 1, 40~50 \uc0ac\uc774\uc5d0 \uc788\uc73c\uba74 2, 50~60 \uc0ac\uc774\uc5d0 \uc788\uc73c\uba74 3 \uc774\ub807\uac8c \ub123\uc5b4\ubcf4\uaca0\uc2b5\ub2c8\ub2e4. <\/p>\n\n\n\n
\ubc94\uc8fc\ud615 : \uba87\uac1c\uc758 \ubc94\uc8fc\ub85c \ub098\ub220\uc9c4 \uc790\ub8cc\ub85c \uba85\ubaa9\ud615, \uc21c\uc11c\ud615 \ub370\uc774\ud130\uac00 \uc788\uc74c
* \uba85\ubaa9\ud615 : \uc131\ubcc4, \uc131\uacf5\uc5ec\ubd80, \ud608\uc561\ud615 \ub4f1 \ub2e8\uc21c\ud788 \ubd84\ub958\ub41c \uc790\ub8cc
* \uc21c\uc11c\ud615 : \uac1c\uac1c\uc758 \uac12\ub4e4\uc774 \uc774\uc0b0\uc801\uc774\uba70 \uadf8\ub4e4 \uc0ac\uc774\uc5d0 \uc21c\uc11c \uad00\uacc4\uac00 \uc874\uc7ac\ud558\ub294 \uc790\ub8cc<\/p>\n\n\n\n
\uc218\uce58\ud615 : \uc774\uc0b0\ud615\uacfc \uc5f0\uc18d\ud615\uc73c\ub85c \uc774\ub904\uc9c4 \uc790\ub8cc\ub85c \uc774\uc0b0\ud615\uacfc \uc5f0\uc18d\uc801\uc778 \ud615\ud0dc\uc758 \uc790\ub8cc\uac00 \uc788\uc74c
* \uc774\uc0b0\ud615 : \uc774\uc0b0\uc801\uc778 \uac12\uc744 \uac16\ub294 \ub370\uc774\ud130\ub85c \ucd9c\uc0b0 \ud69f\uc218 \ub4f1\uc744 \uc758\ubbf8
* \uc5f0\uc18d\ud615 : \uc5f0\uc18d\uc801\uc778 \uac12\uc744 \uac16\ub294 \ub370\uc774\ud130\ub85c \ud0a4\ub098 \ubab8\ubb34\uac8c \ub4f1\uc744 \uc758\ubbf8<\/p>\n\n\n\n
dataset.loc[ dataset['age'] < 40, 'ACODE']= 1\ndataset.loc[ (dataset['age'] >= 40) & (dataset['age'] <50), 'ACODE']= 2\ndataset.loc[ (dataset['age'] >= 50) & (dataset['age'] <60), 'ACODE']= 3<\/pre>\n\n\n\n
\ud6c8\ub828\uc5d0 \uc0ac\uc6a9\ud558\ub294 \uceec\ub7fc\uc740 \ub098\uc774\ub97c \uc81c\uc678\ud55c \ub098\uba38\uc9c0 \uceec\ub7fc\ub4e4\uc785\ub2c8\ub2e4. \uc5b4\ub5a4 \ub0b4\uc6a9\ub4e4\uc774 \uc788\ub294\uc9c0 \uc0b4\ud3b4\ubcf4\uba74 \uc544\ub798\uc640 \uac19\uc2b5\ub2c8\ub2e4. <\/p>\n\n\n\n
\"\"<\/figure>\n\n\n\n
\ud574\ub2f9 \ub370\uc774\ud130\ub97c 3\ucc28\uc6d0 \ud3c9\uba74\uc5d0 \ub098\ud0c0\ub0b4\ubcf4\uba74 \uc544\ub798\uc640 \uac19\uc2b5\ub2c8\ub2e4. \uadf8\ub798\ud504\ub97c \ubcf4\uba74 3\uac1c\uc758 \ud3ec\uc778\ud130\uac00 \ub2e4\ub978 \uc0c9\uc73c\ub85c \ubd84\ub958\ub418\uc5b4 \uc788\uc2b5\ub2c8\ub2e4. \ub370\uc774\ud130\uac00 \ub9ce\uc9c0 \uc54a\uace0 5\ucc28\uc6d0 \ub370\uc774\ud130\ub97c 3\ucc28\uc6d0\uc73c\ub85c \ucd95\uc18c\ud55c \ud615\ud0dc\uc774\uae30 \ub54c\ubb38\uc5d0 \uac01 \ubd84\ub958\uac00 \uc815\ud655\ud558\uc9c0 \uc54a\uc740\uac83 \uac19\uc2b5\ub2c8\ub2e4. \ud559\uc2b5 \ub370\uc774\ud130\uac00 \uc88b\uc740 \ud615\ud0dc\ub85c \ub418\uc5b4 \uc788\ub2e4\uba74 \uc544\ub9c8\ub3c4 \uadf8\ub798\ud504\uc758 \uac01 \uc694\uc18c\ub4e4\uc774 \uc798 \uad6c\ubd84\ub418\uc5b4 \uc788\uc744\ud14c\uc9c0\ub9cc \uc544\uc27d\uac8c\ub3c4 \uc608\uc81c \ub370\uc774\ud130\ub294 \uadf8\ub807\uc9c0 \uc54a\uc740\ub4ef\ud569\ub2c8\ub2e4.<\/p>\n\n\n\n
from mpl_toolkits.mplot3d import Axes3D\n\nfig = plt.figure(1, figsize=(8, 6))\nax = Axes3D(fig, elev=-150, azim=110)\nX_reduced = PCA(n_components=3).fit_transform(dataset[dataset.columns[:-1]])\nax.scatter(X_reduced[:, 0], X_reduced[:, 1], X_reduced[:, 2], c=y,\n           cmap=plt.cm.Set1, edgecolor='k', s=40)\nax.set_title(\"First three PCA directions\")\nax.set_xlabel(\"1st eigenvector\")\nax.w_xaxis.set_ticklabels([])\nax.set_ylabel(\"2nd eigenvector\")\nax.w_yaxis.set_ticklabels([])\nax.set_zlabel(\"3rd eigenvector\")\nax.w_zaxis.set_ticklabels([])\n\nplt.show()<\/pre>\n\n\n\n
\"\"<\/figure>\n\n\n\n
\uc774\uc81c sklearn \ud328\ud0a4\uc9c0\uc758 \ubd84\ub958\ubb38\uc81c\ub97c \ud480\uc218 \uc788\ub294 LogisticRegression\ub97c \uc0ac\uc6a9\ud574\ubcf4\uaca0\uc2b5\ub2c8\ub2e4. \uc0ac\uc6a9\ud558\ub294 \ubc29\ubc95\uc740 LinearRegression\uacfc \uac19\uc774 \uac04\ub2e8\ud558\uac8c \uc0ac\uc6a9\ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4. x \ub370\uc774\ud130\uc640 y \ub370\uc774\ud130\ub294 \uc704\uc5d0\uc11c \uc124\uba85\ud574\ub4dc\ub9b0 \ub370\uc774\ud130\ub97c \uc0ac\uc6a9\ud569\ub2c8\ub2e4.<\/p>\n\n\n\n
\uadf8\ub7f0\ub2e4\uc74c LogisticRegression\uc5d0\uc11c \uc635\uc158\uc744 \uc120\ud0dd\ud558\uace0 \ud574\ub2f9 \ubaa8\ub378\uc744 \ud559\uc2b5\ud569\ub2c8\ub2e4. sklearn\uc740 \ud6a8\uc728\uc801\uc778 \ud559\uc2b5\uc744 \uc704\ud574 \uba87\uac00\uc9c0 \ud6a8\uc728\uc801\uc778 \ud559\uc2b5 \uc54c\uace0\ub9ac\uc998\uc744 \uc81c\uc2dc\ud569\ub2c8\ub2e4. \ud2b9\ud788 feature\uc758 \uc218, \ud559\uc2b5 \ub370\uc774\ud130\uc758 \uc591, \ubd84\ub958\ud560 \ubb38\uc81c\uc758 \uc218 \ub4f1\uc744 \ub530\ub77c\uc11c \ud544\uc694\ud55c \uc54c\uace0\ub9ac\uc998\uc744 \uc120\ud0dd\ud569\ub2c8\ub2e4.<\/p>\n\n\n\n
\uc774\ubc88 \uc608\uc81c\ub294 3\uac1c\uc758 \ubd84\ub958 \ubb38\uc81c(multiclass problems)\ub97c \ud480\uc5b4\uc57c \ud558\uae30 \ub54c\ubb38\uc5d0 ‘newton-cg’ \uc54c\uace0\ub9ac\uc998\uc744 \uc0ac\uc6a9\ud558\ub3c4\ub85d \ud558\uaca0\uc2b5\ub2c8\ub2e4. \uc774 \uc678\uc5d0\ub3c4 \uc5ec\ub7ec \uc54c\uace0\ub9ac\uc998\uacfc \ud30c\ub77c\uba54\ud130\uac00 \uc788\uc2b5\ub2c8\ub2e4. sklearn \uacf5\uc2dd \uc0ac\uc774\ud2b8\uc5d0\uc11c \ud574\ub2f9 \ub0b4\uc6a9\ub4e4\uc744 \ud655\uc774\ud574\ubcf4\uc2dc\uae30 \ubc14\ub78d\ub2c8\ub2e4.<\/p>\n\n\n\n
\"\"
https:\/\/scikit-learn.org\/stable\/modules\/generated\/sklearn.linear_model.LogisticRegression.html<\/figcaption><\/figure>\n\n\n\n
x = dataset[dataset.columns[1:-1]].values\ny = dataset['ACODE'].values\n\nfrom sklearn.linear_model import LogisticRegression\nclf = LogisticRegression(solver='newton-cg', max_iter=100).fit(x, y)\n\npredict = clf.predict(x)\naccuracy=(predict.reshape(-1,1) == y.reshape(-1,1)).sum()\/len(y)\nprint(accuracy)\n# 0.8387096774193549 \uc815\ud655\ub3c4<\/pre>\n\n\n\n
\uc54c\uace0\ub9ac\uc998\uc744 \ud1b5\ud574 \ud655\uc778\ud574\ubcf4\ub2c8 0.8387 \uc758 \uc815\ud655\ub3c4\ub97c \ubcf4\uc5ec\uc90d\ub2c8\ub2e4. \ud559\uc2b5\uc6a9 \ub370\uc774\ud130\uac00 \uc791\uae30 \ub54c\ubb38\uc5d0 \ud14c\uc2a4\ud2b8\uc6a9 \ub370\uc774\ud130\ub97c \ub9cc\ub4e4 \uc218 \uc5c6\uc5c8\ub358 \uc774\uc720\ub85c \uc815\ud655\ub3c4 \uacc4\uc0b0\uc744 \ud55c\ub2e4\uace0 \ud558\uae30\ub3c4 \uc880 \uc560\ub9e4\ud558\uc9c0\ub9cc \ub354 \ub9ce\uc740 \ub370\uc774\ud130\ub97c \ud1b5\ud574\uc11c \ud6c8\ub828\uacfc \ud14c\uc2a4\ud2b8\ub97c \ud574\ubcf4\uc2dc\uae38 \uad8c\ud574\ub4dc\ub9bd\ub2c8\ub2e4.<\/p>\n\n\n\n
\uadf8\ub9ac\uace0 \ud559\uc2b5\uc774 \uc644\ub8cc\ub41c \ud6c4\uc5d0 \uae30\uc6b8\uae30\uc640 \uc808\ud3b8 \uc815\ubcf4\ub97c \ucd9c\ub825\ud574\ubcf4\uba74 \uc544\ub798\uc640 \uac19\uc2b5\ub2c8\ub2e4.<\/p>\n\n\n\n
print('W:{}, b:{}'.format(clf.coef_.T, clf.intercept_.T))\nW:[[ 0.46533128 -0.20477993 -0.26055162]\n [ 0.60397447 -0.08600717 -0.5179662 ]\n [-0.13946355  0.42710757 -0.28764394]\n [ 0.21778426 -0.00474545 -0.2130358 ]\n [ 0.00648806  0.03361824 -0.04010692]\n [ 0.05094977 -0.07150677  0.02055985]], b:[-116.45602409   30.69276337   85.76326072]<\/pre>\n\n\n\n
\uadf8\ub9ac\uace0 \uc774\ub7ec\ud55c \ub370\uc774\ud130\ub97c \ud1b5\ud574\uc11c \uc9c1\uc811 \ud589\ub82c\uacf1\uc744 \ud574\ubd10\ub3c4 \uac19\uc740 \uacb0\uacfc\uac00 \ub098\uc624\ub294 \uac83\uc744 \ud655\uc778\ud560 \uc218 \uc788\uc2b5\ub2c8\ub2e4.<\/p>\n\n\n\n
y_hat = np.matmul(x,clf.coef_.T)+clf.intercept_\ny_hat.argmax(axis=1)+1<\/pre>\n\n\n\n
<\/p>\n","protected":false},"excerpt":{"rendered":"
\uc774\uc804\uae4c\uc9c0\ub294 age, weight, oxygen, runtime, runpulse, rstpulse \ub370\uc774\ud130\ub97c \ud1b5\ud574\uc11c maxpulse\ub97c \uc608\uce21\ud558\ub294 \uc608\uc81c\ub97c \uc0b4\ud3b4\ubd24\uc2b5\ub2c8\ub2e4. maxpulse\ub294 \uc5f0\uc18d\ub41c \uac12\uc744 \uac00\uc9c0\uace0 \uc788\ub294 \uc218\uce58-\uc5f0\uc18d\ud615 \ub370\uc774\ud130\uc785\ub2c8\ub2e4. \ub370\uc774\ud130\uc758 \ud2b9\uc9d5\uc740 \uc544\ub798\uc640 \uac19\uc2b5\ub2c8\ub2e4. RangeIndex: 31 entries, 0 to 30Data columns (total 7 columns):# Column Non-Null Count Dtype— —— ————– —–0 age 31 non-null int641 weight 31 non-null float642 oxygen 31 non-null float643 runtime 31 … <\/p>\n
\ub354 \ubcf4\uae30 “\ub85c\uc9c0\uc2a4\ud2f1 \ud68c\uadc0(Logistic Regression) \ub2e4\uc911\ubd84\ub958”<\/span><\/a><\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":[],"categories":[17,70],"tags":[54,48,57,45,56,55],"_links":{"self":[{"href":"http:\/\/blog.cedartrees.co.kr\/index.php\/wp-json\/wp\/v2\/posts\/656"}],"collection":[{"href":"http:\/\/blog.cedartrees.co.kr\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"http:\/\/blog.cedartrees.co.kr\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"http:\/\/blog.cedartrees.co.kr\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"http:\/\/blog.cedartrees.co.kr\/index.php\/wp-json\/wp\/v2\/comments?post=656"}],"version-history":[{"count":10,"href":"http:\/\/blog.cedartrees.co.kr\/index.php\/wp-json\/wp\/v2\/posts\/656\/revisions"}],"predecessor-version":[{"id":775,"href":"http:\/\/blog.cedartrees.co.kr\/index.php\/wp-json\/wp\/v2\/posts\/656\/revisions\/775"}],"wp:attachment":[{"href":"http:\/\/blog.cedartrees.co.kr\/index.php\/wp-json\/wp\/v2\/media?parent=656"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"http:\/\/blog.cedartrees.co.kr\/index.php\/wp-json\/wp\/v2\/categories?post=656"},{"taxonomy":"post_tag","embeddable":true,"href":"http:\/\/blog.cedartrees.co.kr\/index.php\/wp-json\/wp\/v2\/tags?post=656"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}