\n
import pandas as pd<\/pre>\n\n\n\nimport numpy as np<\/pre>\n\n\n\nimport matplotlib.pyplot as plt<\/pre>\n\n\n\nfrom sklearn.cluster import KMeans<\/pre>\n\n\n\n%matplotlib inline<\/pre>\n<\/div><\/div>\n\n\n\nNh\u01b0 b\u1ea1n c\u00f3 th\u1ec3 th\u1ea5y t\u1eeb \u0111o\u1ea1n code tr\u00ean, ch\u00fang ta s\u1ebd nh\u1eadp c\u00e1c th\u01b0 vi\u1ec7n sau v\u00e0o d\u1ef1 \u00e1n c\u1ee7a m\u00ecnh:<\/p>\n\n\n\n
Panda \u0111\u1ec3 \u0111\u1ecdc v\u00e0 vi\u1ebft b\u1ea3ng t\u00ednh<\/p>\n\n\n\n
Numpy \u0111\u1ec3 th\u1ef1c hi\u1ec7n c\u00e1c ph\u00e9p t\u00ednh hi\u1ec7u qu\u1ea3<\/p>\n\n\n\n
Matplotlib \u0111\u1ec3 tr\u1ef1c quan h\u00f3a d\u1eef li\u1ec7u<\/p>\n\n\n\n
B\u01b0\u1edbc 2:<\/strong> t\u1ea1o d\u1eef li\u1ec7u ng\u1eabu nhi\u00ean<\/p>\n\n\n\nCh\u00fang ta s\u1ebd nh\u1eadp \u0111o\u1ea1n code sau \u0111\u1ec3 t\u1ea1o m\u1ed9t s\u1ed1 d\u1eef li\u1ec7u ng\u1eabu nhi\u00ean trong kh\u00f4ng gian 2 chi\u1ec1u v\u00e0 th\u1ec3 hi\u1ec7n c\u00e1c d\u1eef li\u1ec7u \u0111\u00f3 tr\u00ean bi\u1ec3u \u0111\u1ed3 scatter plot:<\/p>\n\n\n\n
X= -2 * np.random.rand(200,2)<\/pre>\n\n\n\nX1 = 1 + 2 * np.random.rand(100,2)<\/pre>\n\n\n\nX[100:200, :] = X1<\/pre>\n\n\n\nplt.scatter(X[ : , 0], X[ :, 1], s = 50, c = 'b')<\/pre>\n\n\n\nplt.show()<\/pre>\n\n\n\nT\u1ed5ng c\u1ed9ng 200 \u0111i\u1ec3m d\u1eef li\u1ec7u \u0111\u00e3 \u0111\u01b0\u1ee3c t\u1ea1o v\u00e0 chia th\u00e0nh hai nh\u00f3m, m\u1ed7i nh\u00f3m 100 \u0111i\u1ec3m.<\/p>\n\n\n\n
C\u00e1c d\u1eef li\u1ec7u n\u00e0y \u0111\u01b0\u1ee3c hi\u1ec3n th\u1ecb tr\u00ean kh\u00f4ng gian hai chi\u1ec1u nh\u01b0 sau:<\/p>\n\n\n\n
![\"\"](\"https:\/\/i0.wp.com\/datasciencebasic.com\/wp-content\/uploads\/2021\/04\/scatter-plot.png?resize=384%2C262&ssl=1\")
<\/figure><\/div>\n\n\n\nB\u01b0\u1edbc 3: <\/strong>S\u1eed d\u1ee5ng Scikit-Learn<\/p>\n\n\n\nCh\u00fang ta s\u1ebd s\u1eed d\u1ee5ng ch\u1ee9c n\u0103ng K-means c\u00f3 s\u1eb5n trong th\u01b0 vi\u1ec7n Scikit-learning \u0111\u1ec3 x\u1eed l\u00fd d\u1eef li\u1ec7u ng\u1eabu nhi\u00ean v\u1eeba \u0111\u01b0\u1ee3c t\u1ea1o \u1edf tr\u00ean.<\/p>\n\n\n\n
Nh\u1eadp code nh\u01b0 sau:<\/p>\n\n\n\n
from sklearn.cluster import KMeans<\/pre>\n\n\n\nKmean = KMeans(n_clusters=2)<\/pre>\n\n\n\nKmean.fit(X)<\/pre>\n\n\n\nTrong tr\u01b0\u1eddng h\u1ee3p n\u00e0y, ch\u00fang ta \u0111\u00e3 gi\u1ea3 \u0111\u1ecbnh tr\u01b0\u1edbc s\u1ed1 c\u1ee5m hay k (n_clusters) l\u00e0 2.<\/p>\n\n\n\n
B\u01b0\u1edbc 4:<\/strong> T\u00ecm centroid<\/p>\n\n\n\n\u0110\u00e2y l\u00e0 code \u0111\u1ec3 t\u00ecm \u0111i\u1ec3m trung t\u00e2m (centroid) c\u1ee7a c\u00e1c c\u1ee5m:<\/p>\n\n\n\n
Kmean.cluster_centers_<\/pre>\n\n\n\nK\u1ebft qu\u1ea3 c\u1ee7a gi\u00e1 tr\u1ecb c\u1ee7a centroid trong tr\u01b0\u1eddng h\u1ee3p n\u00e0y l\u00e0:<\/p>\n\n\n\n
array([[ 1.87977152, 1.98393029],\n [-1.03261149, -0.92568723]])<\/pre>\n\n\n\nCh\u00fang ta c\u00f3 th\u1ec3 th\u1ec3 hi\u1ec7n c\u00e1c centroid n\u00e0y tr\u00ean s\u01a1 \u0111\u1ed3 scatter plot c\u00f9ng v\u1edbi c\u00e1c \u0111i\u1ec3m d\u1eef li\u1ec7u ng\u1eabu nhi\u00ean, s\u1eed d\u1ee5ng code nh\u01b0 sau:<\/p>\n\n\n\n
plt.scatter(X[ : , 0], X[ : , 1], s =50, c='b')<\/pre>\n\n\n\nplt.scatter( 1.87977152,1.98393029, s=100, c='g', marker='s')<\/pre>\n\n\n\nplt.scatter(-1.03261149, -0.92568723, s=100, c='r', marker='s')<\/pre>\n\n\n\nplt.show()<\/pre>\n\n\n\nK\u1ebft qu\u1ea3 thu \u0111\u01b0\u1ee3c:<\/p>\n\n\n\n
![\"\"](\"https:\/\/i0.wp.com\/datasciencebasic.com\/wp-content\/uploads\/2021\/04\/scatter-plot-2.png?resize=426%2C253&ssl=1\")
<\/figure><\/div>\n\n\n\nB\u01b0\u1edbc 5:<\/strong> Ki\u1ec3m tra k\u1ebft qu\u1ea3 thu\u1eadt to\u00e1n<\/p>\n\n\n\nS\u1eed d\u1ee5ng code nh\u01b0 sau \u0111\u1ec3 l\u1ea5y thu\u1ed9c t\u00ednh nh\u00e3n c\u1ee7a t\u1eadp d\u1eef li\u1ec7u v\u1eeba \u0111\u01b0\u1ee3c s\u1eed d\u1ee5ng \u0111\u1ec3 ph\u00e2n c\u1ee5m K-means; ngh\u0129a l\u00e0, c\u00e1ch c\u00e1c \u0111i\u1ec3m d\u1eef li\u1ec7u \u0111\u01b0\u1ee3c ph\u00e2n lo\u1ea1i th\u00e0nh hai c\u1ee5m.<\/p>\n\n\n\n
Kmean.labels_<\/pre>\n\n\n\nK\u1ebft qu\u1ea3:<\/p>\n\n\n\n
array([1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,<\/pre>\n\n\n\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,<\/pre>\n\n\n\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,<\/pre>\n\n\n\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,<\/pre>\n\n\n\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,<\/pre>\n\n\n\n 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,<\/pre>\n\n\n\n 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,<\/pre>\n\n\n\n 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,<\/pre>\n\n\n\n 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,<\/pre>\n\n\n\n 0, 0], dtype=int32)<\/pre>\n\n\n\nNh\u01b0 b\u1ea1n c\u00f3 th\u1ec3 th\u1ea5y \u1edf tr\u00ean, 100 \u0111i\u1ec3m d\u1eef li\u1ec7u thu\u1ed9c v\u1ec1 0 cluster trong khi ph\u1ea7n c\u00f2n l\u1ea1i thu\u1ed9c v\u1ec1 1 cluster.<\/p>\n\n\n\n
Ti\u1ebfp theo, ch\u1eb3ng h\u1ea1n ch\u00fang ta c\u00f3 m\u1ed9t \u0111i\u1ec3m d\u1eef li\u1ec7u b\u1ea5t k\u1ef3, ta c\u1ea7n ki\u1ec3m tra xem \u0111i\u1ec3m d\u1eef li\u1ec7u n\u00e0y thu\u1ed9c c\u1ee5m n\u00e0o. H\u00e3y s\u1eed d\u1ee5ng code d\u01b0\u1edbi \u0111\u00e2y:<\/p>\n\n\n\n
sample_test=np.array([-3.0,-3.0])<\/pre>\n\n\n\nsecond_test=sample_test.reshape(1, -1)<\/pre>\n\n\n\nKmean.predict(second_test)<\/pre>\n\n\n\n\u0110\u00e2y l\u00e0 k\u1ebft qu\u1ea3:<\/p>\n\n\n\n
array([0], dtype=int32)<\/pre>\n\n\n\nNh\u01b0 v\u1eady d\u1eef li\u1ec7u ch\u00fang ta v\u1eeba ki\u1ec3m tra thu\u1ed9c v\u1ec1 c\u1ee5m 0 (t\u00e2m m\u00e0u xanh l\u00e1 c\u00e2y).<\/p>\n\n\n\n
K\u1ebft lu\u1eadn<\/strong><\/p>\n\n\n\nK-means clustering l\u00e0 m\u1ed9t k\u1ef9 thu\u1eadt \u0111\u01b0\u1ee3c s\u1eed d\u1ee5ng r\u1ed9ng r\u00e3i \u0111\u1ec3 ph\u00e2n t\u00edch c\u1ee5m d\u1eef li\u1ec7u.<\/p>\n\n\n\n
Ph\u01b0\u01a1ng ph\u00e1p n\u00e0y r\u1ea5t d\u1ec5 hi\u1ec3u v\u00e0 d\u1ec5 th\u1ef1c hi\u1ec7n. H\u01a1n n\u1eefa, s\u1eed d\u1ee5ng ph\u01b0\u01a1ng ph\u00e1p ph\u00e2n c\u1ee5m K-means cho k\u1ebft qu\u1ea3 \u0111\u00e0o t\u1ea1o nhanh. Tuy nhi\u00ean, hi\u1ec7u su\u1ea5t c\u1ee7a n\u00f3 th\u01b0\u1eddng kh\u00f4ng t\u1ed1t b\u1eb1ng c\u00e1c k\u1ef9 thu\u1eadt ph\u00e2n c\u1ee5m ph\u1ee9c t\u1ea1p kh\u00e1c b\u1edfi v\u00ec c\u00e1c bi\u1ebfn th\u1ec3 nh\u1ecf trong d\u1eef li\u1ec7u c\u00f3 th\u1ec3 d\u1eabn \u0111\u1ebfn sai s\u1ed1 l\u1edbn.<\/p>\n\n\n\n
H\u01a1n n\u1eefa, c\u00e1c c\u1ee5m \u0111\u01b0\u1ee3c gi\u1ea3 \u0111\u1ecbnh l\u00e0 h\u00ecnh c\u1ea7u v\u00e0 c\u00f3 k\u00edch th\u01b0\u1edbc \u0111\u1ed3ng \u0111\u1ec1u, \u0111i\u1ec1u n\u00e0y c\u00f3 th\u1ec3 l\u00e0m gi\u1ea3m \u0111\u1ed9 ch\u00ednh x\u00e1c c\u1ee7a c\u00e1c k\u1ebft qu\u1ea3 ph\u00e2n c\u1ee5m.<\/p>\n\n\n\n
Trong c\u00e1c b\u00e0i vi\u1ebft ti\u1ebfp theo, t\u00f4i s\u1ebd gi\u1edbi thi\u1ec7u c\u00e1c ph\u01b0\u01a1ng ph\u00e1p ph\u00e2n c\u1ee5m kh\u00e1c nh\u1eb1m kh\u1eafc ph\u1ee5c nh\u1eefng \u0111i\u1ec3m y\u1ebfu c\u1ee7a ph\u01b0\u01a1ng ph\u00e1p ph\u00e2n c\u1ee5m K-means.<\/p>\n","protected":false},"excerpt":{"rendered":"
K-means clustering l\u00e0 m\u1ed9t trong nh\u1eefng thu\u1eadt to\u00e1n h\u1ecdc m\u00e1y kh\u00f4ng gi\u00e1m s\u00e1t (unsupervised learning) \u0111\u01a1n gi\u1ea3n v\u00e0 ph\u1ed5 bi\u1ebfn. […]<\/p>\n","protected":false},"author":2,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"nf_dc_page":"","om_disable_all_campaigns":false,"_monsterinsights_skip_tracking":false,"_jetpack_newsletter_access":"","_jetpack_dont_email_post_to_subs":false,"_jetpack_newsletter_tier_id":0,"_jetpack_memberships_contains_paywalled_content":false,"_jetpack_memberships_contains_paid_content":false,"footnotes":""},"categories":[3],"tags":[],"class_list":["post-126","post","type-post","status-publish","format-standard","hentry","category-kien-thuc"],"jetpack_sharing_enabled":true,"jetpack_featured_media_url":"","_links":{"self":[{"href":"https:\/\/datasciencebasic.com\/index.php?rest_route=\/wp\/v2\/posts\/126","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/datasciencebasic.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/datasciencebasic.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/datasciencebasic.com\/index.php?rest_route=\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/datasciencebasic.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=126"}],"version-history":[{"count":2,"href":"https:\/\/datasciencebasic.com\/index.php?rest_route=\/wp\/v2\/posts\/126\/revisions"}],"predecessor-version":[{"id":147,"href":"https:\/\/datasciencebasic.com\/index.php?rest_route=\/wp\/v2\/posts\/126\/revisions\/147"}],"wp:attachment":[{"href":"https:\/\/datasciencebasic.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=126"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/datasciencebasic.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=126"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/datasciencebasic.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=126"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}
![\"\"](\"https:\/\/i0.wp.com\/datasciencebasic.com\/wp-content\/uploads\/2021\/04\/centroid.png?resize=695%2C283&ssl=1\")
<\/figure><\/div>\n\n\n\n