{"id":471,"date":"2019-07-22T20:20:00","date_gmt":"2019-07-22T12:20:00","guid":{"rendered":"http:\/\/note.systw.net\/note\/?p=471"},"modified":"2023-11-02T20:22:57","modified_gmt":"2023-11-02T12:22:57","slug":"sklearn-feature","status":"publish","type":"post","link":"https:\/\/systw.net\/note\/archives\/471","title":{"rendered":"SKLearn Feature"},"content":{"rendered":"\n

\u7279\u5fb5\u6e1b\u5c11\u6700\u5e38\u7684\u5169\u7a2e\u65b9\u5f0f
\u7279\u5fb5\u58d3\u7e2e(\u964d\u7dad)<\/strong>\uff1a\u900f\u904e\u6f14\u7b97\u6cd5\u5c07\u591a\u500b\u7279\u5fb5\u58d3\u7e2e\u6210\u5c11\u91cf\u7279\u5fb5
\u7279\u5fb5\u6311\u9078<\/strong>\uff1a\u900f\u904e\u6f14\u7b97\u6cd5\u5c07\u6700\u5177\u6709\u4ee3\u8868\u6027\u7684\u7279\u5fb5\u6311\u51fa\u4f86<\/p>\n\n\n\n

………………………………………………………………………………..<\/p>\n\n\n\n

Decomposing(\u7279\u5fb5\u58d3\u7e2e)<\/h2>\n\n\n\n

\u964d\u7dad\u7684\u65b9\u5f0f\u5f88\u591a\uff0c\u50cf\u662fPCA,TruncatedSVD,LDA\u7b49\u7b49\uff0c\u4ee5\u4e0b\u4ecb\u7d39PCA\u7684\u65b9\u6cd5
refer
http:\/\/scikit-learn.org\/stable\/modules\/decomposition.html<\/p>\n\n\n\n

PCA<\/strong>
\u50c5\u964d\u7dad\u4e0d\u6311\u9078\u95dc\u9375\u7dad\u5ea6
ps:
PCA doesn’t eliminate dimensions and keeps others from the original data. It transforms your data in a number of dimensions whose data are completely different from the original ones.
refer
http:\/\/scikit-learn.org\/stable\/modules\/generated\/sklearn.decomposition.PCA.html#sklearn.decomposition.PCA<\/p>\n\n\n\n


\u8f09\u5165\u6a21\u7d44<\/strong>
from sklearn.decomposition import PCA
\u6307\u5b9a\u8981\u964d\u5230\u5e7e\u7dad<\/strong>
pca = PCA(n_components=2)
\u6307\u5b9a\u8981\u964d\u7dad\u7684\u8cc7\u6599<\/strong>
data_pca = pca.fit_transform(data)<\/p>\n\n\n\n


example
\u5c073\u7dad\u7684iris data\u8b8a\u62102\u7dad<\/strong>
>>>from sklearn.decomposition import PCA
>>>pca = PCA(n_components=2)
>>>data_pca = pca.fit_transform(iris.data)
>>>print data_pca.shape
(150, 2)
>>> print(pca.explained_variance_ratio_)
array([ 0.92461621, 0.05301557])<\/p>\n\n\n\n


example
\u5c07iris\u8cc7\u6599\u96c6\u7528PCA\u8b8a\u62102\u7dad\u4e26\u756b\u5728\u4e00\u5f35\u5716\u4e0a<\/strong><\/p>\n\n\n\n

#!\/usr\/bin\/env python\n\nimport pylab as pl\nfrom itertools import cycle\nfrom sklearn.decomposition import PCA\nfrom sklearn import datasets\nimport numpy as np\n\ndef plot_2D(data, target, target_names):\n\u3000colors = cycle('rgbcmykw')\n\u3000target_ids = range(len(target_names))\n\u3000pl.figure()\n\u3000for i, c, label in zip(target_ids, colors, target_names):\n\u3000\u3000pl.scatter(data[target == i, 0], data[target == i, 1],c=c, label=label)\n\u3000pl.legend()\n\u3000pl.savefig(\"pca.png\")\n\niris = datasets.load_iris()\npca = PCA(n_components=2)\nX_pca = pca.fit_transform(iris.data)\n\nnp.round(X_pca.mean(axis=0), decimals=5)\nnp.round(X_pca.std(axis=0), decimals=5)\nnp.round(np.corrcoef(X_pca.T), decimals=5)\n\nplot_2D(X_pca, iris.target, iris.target_names)<\/code><\/pre>\n\n\n\n
<\/p>\n\n\n\n
ps:
pl.scatter
http:\/\/matplotlib.org\/api\/pyplot_api.html#matplotlib.pyplot.scatter

ps:
colors = cycle(‘rgbcmykw’)
b: blue
g: green
r: red
c: cyan
m: magenta
y: yellow
k: black
w: white

ps:
matplotlib.colors
http:\/\/matplotlib.org\/api\/colors_api.html<\/p>\n\n\n\n
……………………………………………………………………………………… <\/p>\n\n\n\n
Feature Selection(\u7279\u5fb5\u6311\u9078)<\/h2>\n\n\n\n
\u7279\u5fb5\u6311\u9078\u7684\u65b9\u5f0f\u6709\u5f88\u591a\u65b9\u6cd5\uff0c\u5e38\u898b\u7684\u6709\u4ee5\u4e0b\u4e09\u7a2e\uff1a
Univariate feature selection\u3000
Recursive feature elimination\u3000
Feature selection using SelectFromModel<\/p>\n\n\n\n
\u4ee5\u4e0b\u4e3b\u8981\u4ecb\u7d39Univariate feature selection\u7684\u65b9\u6cd5\uff0c\u6b64\u65b9\u6cd5\u7684\u539f\u7406\u662f\u5206\u522b\u55ae\u7368\u7684\u8ba1\u7b97\u6bcf\u4e2a\u7279\u5fb5\u7684\u67d0\u500b\u7d71\u8a08\u6307\u6a19\uff0c\u5728\u6839\u64da\u8a72\u6307\u6a19\u5224\u65b7\u91cd\u8981\u6027<\/p>\n\n\n\n
……………………………………………………………………………………<\/p>\n\n\n\n

Univariate feature selection<\/p>\n\n\n\n
\u4e3b\u8981\u6709\u4ee5\u4e0b\u5e7e\u7a2e\u65b9\u6cd5
SelectKBest\uff1a<\/strong>\u6392\u540d\u6392\u5728\u524dn\u500b\u7684\u7279\u5fb5
SelectPercentile\uff1a<\/strong>\u6392\u540d\u6392\u5728\u524dn%\u7684\u7279\u5fb5
\u3000
…<\/p>\n\n\n\n
SelectPercentile(score_func, percentile)<\/strong>
score_func :<\/strong> callable, ex:f_classif
Function taking two arrays X and y, and returning a pair of arrays (scores, pvalues).
percentile : <\/strong>int, optional, default=10
Percent of features to keep.<\/p>\n\n\n\n
SelectKBest(score_func,k)<\/strong>
score_func <\/strong>: callable, ex:chi2
Function taking two arrays X and y, and returning a pair of arrays (scores, pvalues).
k <\/strong>: int or “all”, optional, default=10
Number of top features to select. The “all” option bypasses selection, for use in a parameter search.<\/p>\n\n\n\n
score_func\u7684\u9078\u64c7<\/strong>
\u3000\u5c0d\u65bcregression\u554f\u984c\uff0c\u53ef\u4ee5\u7528f_regression
\u3000\u5c0d\u65bcclassification\u554f\u984c\uff0c\u53ef\u4ee5\u7528chi2\u6216\u8005f_classif
ex:
from sklearn.feature_selection import SelectPercentile, f_classif
selector = SelectPercentile(f_classif, percentile=10)<\/p>\n\n\n\n

…<\/p>\n\n\n\n
example by SelectKBest<\/strong><\/p>\n\n\n\n
ex:
classfication\u7684\u7279\u5fb5\u9078\u53d6<\/strong>
#vi univariate.py<\/p>\n\n\n\n
from sklearn import datasets\niris = datasets.load_iris()\nX = iris.data\ny = iris.target\nfrom sklearn.feature_selection import SelectKBest, chi2\nselector = SelectKBest(chi2,k=3)\nresult=selector.fit(X, y)\nprint result.scores_\nx2=selector.transform(X) #\u5f9e\u539f\u59cb\u8cc7\u6599\u96c6\u4e2d\u53d6\u51fa\u524d\u4e09\u597d\u7684\u7279\u5fb5\u505a\u70ba\u65b0\u8cc7\u6599\u96c6\nprint x2<\/code><\/pre>\n\n\n\n
#python univariate.py
[ 10.81782088 3.59449902 116.16984746 67.24482759]
[[ 5.1 1.4 0.2]
[ 4.9 1.4 0.2]
[ 4.7 1.3 0.2]
[ 4.6 1.5 0.2]
…omit…
\u8aaa\u660e
\u7b2c3\u500b\u6b04\u4f4d\u6578\u503c\u6700\u5927\uff0c\u8868\u793a\u6700\u6709\u5f71\u97ff\u529b
\u7b2c2\u500b\u6b04\u4f4d\u6578\u503c\u6700\u5c0f\uff0c\u8868\u793a\u5f71\u97ff\u6548\u679c\u6700\u5dee
[ 10.81782088 3.59449902 116.16984746 67.24482759]<\/p>\n\n\n\n
ex:
regression\u7684\u7279\u5fb5\u9078\u53d6<\/strong>
#vi univariate.py<\/p>\n\n\n\n
from sklearn import datasets\nboston = datasets.load_boston()\nX = boston.data\ny = boston.target\nfrom sklearn.feature_selection import SelectKBest, f_regression\nselector = SelectKBest(f_regression,k=2)\nresult=selector.fit(X, y)\nprint result.scores_\nx2=selector.transform(X)\nprint x2<\/code><\/pre>\n\n\n\n


refer
https:\/\/machine-learning-python.kspax.io\/Feature_Selection\/intro.html
http:\/\/www.cnblogs.com\/jasonfreak\/p\/5448385.html
http:\/\/sklearn.lzjqsdd.com\/modules\/feature_selection.html<\/p>\n","protected":false},"excerpt":{"rendered":"
\u7279\u5fb5\u6e1b\u5c11\u6700\u5e38\u7684\u5169\u7a2e\u65b9\u5f0f\u7279\u5fb5\u58d3\u7e2e(\u964d\u7dad)\uff1a\u900f\u904e\u6f14\u7b97\u6cd5\u5c07\u591a\u500b\u7279\u5fb5 …<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"fifu_image_url":"","fifu_image_alt":"","_jetpack_memberships_contains_paid_content":false,"footnotes":"","jetpack_publicize_message":"","jetpack_publicize_feature_enabled":true,"jetpack_social_post_already_shared":false,"jetpack_social_options":{"image_generator_settings":{"template":"highway","default_image_id":0,"font":"","enabled":false},"version":2}},"categories":[13],"tags":[],"class_list":["post-471","post","type-post","status-publish","format-standard","hentry","category-dataanalysis"],"jetpack_publicize_connections":[],"jetpack_featured_media_url":"","jetpack_sharing_enabled":true,"_links":{"self":[{"href":"https:\/\/systw.net\/note\/wp-json\/wp\/v2\/posts\/471","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/systw.net\/note\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/systw.net\/note\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/systw.net\/note\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/systw.net\/note\/wp-json\/wp\/v2\/comments?post=471"}],"version-history":[{"count":0,"href":"https:\/\/systw.net\/note\/wp-json\/wp\/v2\/posts\/471\/revisions"}],"wp:attachment":[{"href":"https:\/\/systw.net\/note\/wp-json\/wp\/v2\/media?parent=471"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/systw.net\/note\/wp-json\/wp\/v2\/categories?post=471"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/systw.net\/note\/wp-json\/wp\/v2\/tags?post=471"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}