{"id":349,"date":"2017-02-17T14:40:12","date_gmt":"2017-02-17T06:40:12","guid":{"rendered":"http:\/\/www.mrtblog.cn\/?p=349"},"modified":"2023-03-05T14:40:54","modified_gmt":"2023-03-05T06:40:54","slug":"%e6%9c%b4%e7%b4%a0%e8%b4%9d%e5%8f%b6%e6%96%af%e7%ae%97%e6%b3%95%e5%8f%8a%e5%ae%9e%e8%b7%b5%ef%bc%881%ef%bc%89","status":"publish","type":"post","link":"http:\/\/www.mrtblog.cn\/?p=349","title":{"rendered":"\u6734\u7d20\u8d1d\u53f6\u65af\u7b97\u6cd5\u53ca\u5b9e\u8df5"},"content":{"rendered":"<div class='epvc-post-count'><span class='epvc-eye'><\/span>  <span class=\"epvc-count\"> 1,345<\/span><span class='epvc-label'> Views<\/span><\/div><p><em>\u53c2\u8003\u4e66\u7c4d\u300a\u673a\u5668\u5b66\u4e60\u5b9e\u6218\u300b<\/em><\/p>\n<p>KNN\u548c\u51b3\u7b56\u6811\u7b97\u6cd5\u8981\u6c42\u5206\u7c7b\u5668\u505a\u51fa\u8270\u96be\u51b3\u7b56\uff0c\u7ed9\u51fa\u201c\u8be5\u6570\u636e\u5b9e\u4f8b\u5c5e\u4e8e\u54ea\u4e00\u7c7b\u201d\u8fd9\u7c7b\u95ee\u9898\u7684\u660e\u786e\u7b54\u6848\uff0c\u4f46\u6709\u65f6\u5206\u7c7b\u5668\u4f1a\u4ea7\u751f\u9519\u8bef\u7ed3\u679c\u3002\u8fd9\u662f\u53ef\u4ee5\u8981\u6c42\u5206\u7c7b\u5668\u7ed9\u51fa\u4e00\u4e2a\u6700\u4f18\u7684\u7c7b\u522b\u731c\u6d4b\u7ed3\u679c\uff0c\u540c\u65f6\u7ed9\u51fa\u8fd9\u4e2a\u731c\u6d4b\u7684\u6982\u7387\u4f30\u8ba1\u503c\u3002<\/p>\n<h2>\u57fa\u4e8e\u8d1d\u53f6\u65af\u51b3\u7b56\u7406\u8bba\u7684\u5206\u7c7b\u65b9\u6cd5<\/h2>\n<p><strong>\u6734\u7d20\u8d1d\u53f6\u65af<\/strong><br \/>\n\u4f18\u70b9\uff1a\u5728\u6570\u636e\u8f83\u5c11\u7684\u60c5\u51b5\u4e0b\u4ecd\u7136\u6709\u6548\uff0c\u53ef\u4ee5\u5904\u7406\u591a\u7c7b\u522b\u95ee\u9898<br \/>\n\u7f3a\u70b9\uff1a\u5bf9\u4e8e\u8f93\u5165\u6570\u636e\u7684\u51c6\u5907\u65b9\u5f0f\u8f83\u4e3a\u654f\u611f<br \/>\n\u9002\u7528\u6570\u636e\u7c7b\u578b\uff1a\u6807\u79f0\u578b\u6570\u636e<\/p>\n<p>\u5047\u8bbe\u6211\u4eec\u6709\u4e24\u7c7b\u6570\u636e\u96c6\uff0c\u5b83\u7531\u4e24\u7c7b\u6570\u636e\u7ec4\u6210\uff0c\u6570\u636e\u5206\u5e03\u5982\u56fe4-1<br \/>\n<a href=\"http:\/\/www.mrtblog.cn\/wp-content\/uploads\/2017\/02\/Bayes-t4-1.png\"><img loading=\"lazy\" decoding=\"async\" class=\"alignnone size-full wp-image-350\" src=\"http:\/\/www.mrtblog.cn\/wp-content\/uploads\/2017\/02\/Bayes-t4-1.png\" alt=\"\" width=\"555\" height=\"478\" srcset=\"http:\/\/www.mrtblog.cn\/wp-content\/uploads\/2017\/02\/Bayes-t4-1.png 555w, http:\/\/www.mrtblog.cn\/wp-content\/uploads\/2017\/02\/Bayes-t4-1-300x258.png 300w\" sizes=\"auto, (max-width: 555px) 100vw, 555px\" \/><\/a><\/p>\n<p>\u6211\u4eec\u73b0\u5728\u7528p1(x,y)\u8868\u793a\u6570\u636e\u70b9(x,y)\u5c5e\u4e8e\u7c7b\u522b1\uff08\u56fe\u4e2d\u7528\u5706\u70b9\u8868\u793a\u7684\u7c7b\u522b\uff09\u7684\u6982\u7387\uff0c\u7528p2(x,y)\u8868\u793a\u6570\u636e\u70b9(x,y)\u5c5e\u4e8e\u7c7b\u522b2\uff08\u56fe\u4e2d\u7528\u4e09\u89d2\u5f62\u8868\u793a\u7684\u7c7b\u522b\uff09\u7684\u6982\u7387\uff0c\u90a3\u4e48\u5bf9\u4e8e\u4e00\u4e2a\u65b0\u6570\u636e\u70b9(x,y)\uff0c\u53ef\u4ee5\u7528\u4e0b\u9762\u7684\u89c4\u5219\u6765\u5224\u65ad\u5b83\u7684\u7c7b\u522b:<br \/>\np1(x,y)&gt;p2(x,y)\uff0c\u90a3\u4e48\u7c7b\u522b\u4e3a1<br \/>\np2(x,y)&gt;p1(x,y)\uff0c\u90a3\u4e48\u7c7b\u522b\u4e3a2<\/p>\n<p>\u9009\u62e9\u9ad8\u6982\u7387\u5bf9\u5e94\u7684\u7c7b\u522b\uff0c\u8fd9\u5c31\u662f\u8d1d\u53f6\u65af\u51b3\u7b56\u7406\u8bba\u7684\u6838\u5fc3\u3002\u8fd9\u7c7b\u95ee\u9898\u4e0a\uff0c\u4f7f\u7528\u51b3\u7b56\u6811\u6548\u679c\u4e0d\u597d\uff0cKNN\u8ba1\u7b97\u91cf\u592a\u5927\uff0c\u8d1d\u53f6\u65af\u662f\u5f88\u597d\u7684\u9009\u62e9\u3002<\/p>\n<p><strong>\u6761\u4ef6\u6982\u7387\uff1a\u7565<\/strong><br \/>\n<strong>\u8d1d\u53f6\u65af\u516c\u5f0f\uff1a<\/strong><a href=\"http:\/\/www.mrtblog.cn\/wp-content\/uploads\/2017\/02\/Bayes-formula-1.png\"><img loading=\"lazy\" decoding=\"async\" class=\"alignnone size-full wp-image-351\" src=\"http:\/\/www.mrtblog.cn\/wp-content\/uploads\/2017\/02\/Bayes-formula-1.png\" alt=\"\" width=\"217\" height=\"77\"><\/a><\/p>\n<h3>\u4f7f\u7528\u8d1d\u53f6\u65af\u8fdb\u884c\u6587\u6863\u5206\u7c7b<\/h3>\n<p>\u673a\u5668\u5b66\u4e60\u7684\u91cd\u8981\u5e94\u7528\u5c31\u662f\u6587\u6863\u7684\u81ea\u52a8\u5206\u7c7b\u3002\u5728\u6587\u6863\u5206\u7c7b\u4e2d\uff0c\u6574\u4e2a\u6587\u6863\uff08\u5982\u4e00\u5c01\u7535\u5b50\u90ae\u4ef6\uff09\u662f\u5b9e\u4f8b\uff0c\u800c\u7535\u5b50\u90ae\u4ef6\u4e2d\u7684\u67d0\u4e9b\u5143\u7d20\u5219\u6784\u6210\u7279\u5f81\u3002\u867d\u7136\u7535\u5b50\u90ae\u4ef6\u662f\u4e00\u79cd\u4e0d\u65ad\u589e\u52a0\u7684\u6587\u672c\uff0c\u4f46\u6211\u4eec\u540c\u6837\u4e5f\u53ef\u4ee5\u5bf9\u65b0\u95fb\u62a5\u9053\u3001\u7528\u6237\u7559\u8a00\u3001\u653f\u5e9c\u516c\u6587\u7b49\u5176\u5b83\u4efb\u610f\u7c7b\u578b\u7684\u6587\u672c\u8fdb\u884c\u5206\u7c7b\u3002\u6211\u4eec\u53ef\u4ee5\u89c2\u5bdf\u6587\u6863\u4e2d\u51fa\u73b0\u7684\u8bcd\uff0c\u5e76\u628a\u6bcf\u4e2a\u8bcd\u51fa\u73b0\u6216\u8005\u4e0d\u51fa\u73b0\u4f5c\u4e3a\u4e00\u4e2a\u7279\u5f81\uff0c\u8fd9\u6837\u5f97\u5230\u7684\u7279\u5f81\u6570\u76ee\u5c31\u4f1a\u548c\u8bcd\u6c47\u8868\u4e2d\u7684\u8bcd\u76ee\u4e00\u6837\u591a\u3002<\/p>\n<p><strong>\u6734\u7d20\u8d1d\u53f6\u65af\u7684\u4e00\u822c\u8fc7\u7a0b<\/strong><br \/>\n\uff081\uff09\u6536\u96c6\u6570\u636e\uff1a\u53ef\u4ee5\u4f7f\u7528\u4efb\u4f55\u65b9\u6cd5<br \/>\n\uff082\uff09\u51c6\u5907\u6570\u636e\uff1a\u9700\u8981\u6570\u503c\u578b\u6216\u8005\u5e03\u5c14\u578b\u6570\u636e<br \/>\n\uff083\uff09\u5206\u6790\u6570\u636e\uff1a\u6709\u5927\u91cf\u7279\u5f81\u65f6\uff0c\u7ed8\u5236\u7279\u5f81\u4f5c\u7528\u4e0d\u5927\uff0c\u76f4\u65b9\u56fe\u6548\u679c\u66f4\u597d<br \/>\n\uff084\uff09\u8bad\u7ec3\u65b9\u6cd5\uff1a\u8ba1\u7b97\u4e0d\u540c\u7684\u72ec\u7acb\u7279\u5f81\u7684\u6761\u4ef6\u6982\u7387<br \/>\n\uff085\uff09\u6d4b\u8bd5\u7b97\u6cd5\uff1a\u8ba1\u7b97\u9519\u8bef\u7387<br \/>\n\uff086\uff09\u4f7f\u7528\u7b97\u6cd5\uff1a\u4e00\u4e2a\u5e38\u89c1\u7684\u6734\u7d20\u8d1d\u53f6\u65af\u5e94\u7528\u662f\u6587\u6863\u5206\u7c7b\uff0c\u53ef\u4ee5\u5728\u4efb\u610f\u573a\u666f\u4e2d\u4f7f\u7528\u6734\u7d20\u8d1d\u53f6\u65af\u5206\u7c7b\u5668<\/p>\n<p>\u5047\u8bbe\u8bcd\u6c47\u8868\u4e2d\u67091000\u4e2a\u5355\u8bcd\uff0c\u8981\u5f97\u5230\u597d\u7684\u6982\u7387\u5206\u5e03\uff0c\u5c31\u9700\u8981\u8db3\u591f\u591a\u7684\u6570\u636e\u6837\u672c\u3002\u5982\u679c\u6bcf\u4e2a\u7279\u5f81\u9700\u8981N\u4e2a\u6837\u672c\uff0c\u5bf9\u4e8e\u5305\u542b1000\u4e2a\u7279\u5f81\u7684\u8bcd\u6c47\u8868\u5c06\u9700\u8981N^1000\u4e2a\u6837\u672c\u3002\u8bfe\u4ef6\uff0c\u6240\u9700\u6837\u672c\u4f1a\u968f\u7740\u7279\u5f81\u6570\u76ee\u6307\u6570\u589e\u5927\u3002<\/p>\n<p>\u5982\u679c\u7279\u5f81\u76f8\u4e92\u72ec\u7acb\uff0c\u5219\u6837\u672c\u6570\u53ef\u51cf\u5c11\u52301000xN\u3002\u72ec\u7acb\u6307\u4e00\u4e2a\u7279\u5f81\u6216\u8005\u5355\u8bcd\u51fa\u73b0\u7684\u53ef\u80fd\u6027\u4e0e\u5176\u4ed6\u5355\u8bcd\u76f8\u90bb\u6ca1\u6709\u5173\u7cfb\u3002\u4f8b\u5982\u201c\u8ba1\u7b97\u673a\u201d\u548c\u201c\u7a0b\u5e8f\u5458\u201d\u4e00\u8d77\u51fa\u73b0\u7684\u6982\u7387\u6bd4\u201c\u8ba1\u7b97\u673a\u201d\u548c\u201c\u5c0f\u4eba\u4e66\u201d\u7684\u6982\u7387\u5927\uff0c\u4f46\u6211\u4eec\u5047\u8bbe\u4ed6\u4eec\u7684\u6982\u7387\u662f\u76f8\u540c\u7684\uff0c\u5373\u201c\u8ba1\u7b97\u673a\u201d\u3001\u201c\u7a0b\u5e8f\u5458\u201d\u3001\u201c\u5c0f\u4eba\u4e66\u201d\u76f8\u4e92\u72ec\u7acb\uff0c\u8fd9\u4e5f\u662f\u6734\u7d20\u8d1d\u53f6\u65af\u4e2d\u201c\u6734\u7d20\u201d\u7684\u542b\u4e49\u3002\u6734\u7d20\u8d1d\u53f6\u65af\u4e2d\u53e6\u4e00\u4e2a\u5047\u8bbe\u662f\u6bcf\u4e2a\u7279\u5f81\u540c\u7b49\u91cd\u8981\u3002\u867d\u7136\u8fd9\u4e24\u4e2a\u5047\u8bbe\u4e4d\u4e00\u770b\u90fd\u5f88\u6709\u95ee\u9898\uff0c\u4f46\u662f\u5728\u5b9e\u9645\u64cd\u4f5c\u4e2d\u6734\u7d20\u8d1d\u53f6\u65af\u7684\u6548\u679c\u5374\u5f88\u597d\u3002<\/p>\n<hr>\n<h2>\u4f7f\u7528Python\u8fdb\u884c\u6587\u672c\u5206\u7c7b<\/h2>\n<h3>\u51c6\u5907\u6570\u636e\uff1a\u4ece\u6587\u672c\u4e2d\u6784\u5efa\u8bcd\u5411\u91cf<\/h3>\n<p>\u6211\u4eec\u5c06\u628a\u6587\u672c\u770b\u6210\u5355\u8bcd\u5411\u91cf\u6216\u8005\u8bcd\u6761\u5411\u91cf\uff0c\u4e5f\u5c31\u662f\u5c06\u53e5\u5b50\u8f6c\u6362\u6210\u5411\u91cf\u3002\u8003\u8651\u51fa\u73b0\u5728\u6240\u6709\u6587\u6863\u4e2d\u7684\u6240\u6709\u5355\u8bcd\uff0c\u518d\u51b3\u5b9a\u5c06\u54ea\u4e9b\u8bcd\u7eb3\u5165\u8bcd\u6c47\u8868\u3002<\/p>\n<p>\u521b\u5efaBayes.py\u6587\u4ef6\uff0c\u6dfb\u52a0\u4e0b\u5217\u4ee3\u7801\uff1a<\/p>\n<pre class=\"highlight\"><code class=\"language-python line-numbers\">\n#coding=utf-8   \u542b\u6709\u4e2d\u6587\u5fc5\u987b\u89c4\u5b9a\u7f16\u7801\u683c\u5f0f\ndef loadDataSet():#\u521b\u5efa\u8bd5\u9a8c\u6837\u672c\n    postingList = [[&#039;my&#039;,&#039;dog&#039;,&#039;has&#039;,&#039;flea&#039;,&#039;problem&#039;,&#039;help&#039;,&#039;please&#039;],\n        [&#039;maybe&#039;,&#039;not&#039;,&#039;take&#039;,&#039;him&#039;,&#039;to&#039;,&#039;dog&#039;,&#039;park&#039;,&#039;stupid&#039;],\n        [&#039;my&#039;,&#039;dalmation&#039;,&#039;is&#039;,&#039;so&#039;,&#039;cute&#039;,&#039;I&#039;,&#039;love&#039;,&#039;him&#039;],\n        [&#039;stop&#039;,&#039;posting&#039;,&#039;stupid&#039;,&#039;worthless&#039;,&#039;garbage&#039;],\n        [&#039;mr&#039;,&#039;licks&#039;,&#039;ate&#039;,&#039;my&#039;,&#039;steak&#039;,&#039;how&#039;,&#039;to&#039;,&#039;stop&#039;,&#039;him&#039;],\n        [&#039;quit&#039;,&#039;buying&#039;,&#039;worthless&#039;,&#039;dog&#039;,&#039;food&#039;,&#039;stupid&#039;]]\n    classVec = [0,1,0,1,0,1]#0\u4ee3\u8868\u6b63\u5e38\u8a00\u8bba\uff0c1\u8868\u793a\u4fae\u8fb1\u6027\u8bcd\u6c47\n    return postingList,classVec\n\ndef createVocabList(dataSet):#\u7edf\u8ba1\u6240\u6709\u6587\u6863\u4e2d\u51fa\u73b0\u7684\u4e0d\u91cd\u590d\u8bcd\u6c47\n    vocabSet=set([])\n    for document in dataSet:\n        vocabSet = vocabSet | set(document)\n    return list(vocabSet)\n\ndef setOfWords2Vec(vocabList,inputSet):\n    returnVec = [0] * len(vocabList)#\u521b\u5efa\u4e00\u4e2a\u548c\u8f93\u5165\u6587\u6863\u7b49\u957f\u7684\u5411\u91cf\uff0c\u521d\u59cb\u51680\n    for word in inputSet:#\u68c0\u67e5\u8be5\u8bcd\u6c47\u662f\u5426\u51fa\u73b0\u5728\u8bcd\u6c47\u8868\u4e2d\n        if word in vocabList:\n            returnVec[vocabList.index(word)] = 1\n        else:\n            print &quot;the word: %s is not in my Vocabulary!&quot; % word\n    return returnVec#\u5f97\u5230\u5355\u8bcd\u51fa\u73b0\u6b21\u6570\u7684\u5411\u91cf\n<\/code><\/pre>\n<p>\u4e3a\u4e86\u67e5\u770b\u51fd\u6570\u7684\u6267\u884c\u6548\u679c\uff0c\u8f93\u5165\u6d4b\u8bd5\uff1a<\/p>\n<pre class=\"highlight\"><code class=\"language-python line-numbers\">\nimport Bayes\nlistOPosts,listClasses = Bayes.loadDataSet()\nmyVocabList = Bayes.createVocabList(listOPosts)\nprint(myVocabList)\nprint(Bayes.setOfWords2Vec(myVocabList,listOPosts[0]))\nprint(Bayes.setOfWords2Vec(myVocabList,listOPosts[3]))\n<\/code><\/pre>\n<p><a href=\"http:\/\/www.mrtblog.cn\/wp-content\/uploads\/2017\/02\/Bayes-example-1.png\"><img loading=\"lazy\" decoding=\"async\" src=\"http:\/\/www.mrtblog.cn\/wp-content\/uploads\/2017\/02\/Bayes-example-1.png\" alt=\"\" width=\"1390\" height=\"309\" class=\"alignnone size-full wp-image-371\" srcset=\"http:\/\/www.mrtblog.cn\/wp-content\/uploads\/2017\/02\/Bayes-example-1.png 1390w, http:\/\/www.mrtblog.cn\/wp-content\/uploads\/2017\/02\/Bayes-example-1-300x67.png 300w, http:\/\/www.mrtblog.cn\/wp-content\/uploads\/2017\/02\/Bayes-example-1-768x171.png 768w, http:\/\/www.mrtblog.cn\/wp-content\/uploads\/2017\/02\/Bayes-example-1-1024x228.png 1024w\" sizes=\"auto, (max-width: 1390px) 100vw, 1390px\" \/><\/a><\/p>\n<h3>\u8bad\u7ec3\u7b97\u6cd5\uff1a\u4ece\u8bcd\u5411\u91cf\u8ba1\u7b97\u6982\u7387<\/h3>\n<p>\u524d\u9762\u4ecb\u7ecd\u4e86\u5982\u4f55\u5c06\u4e00\u7ec4\u5355\u8bcd\u8f6c\u6362\u4e3a\u4e00\u7ec4\u6570\u5b57\uff0c\u63a5\u4e0b\u6765\u770b\u770b\u5982\u4f55\u4f7f\u7528\u8fd9\u4e9b\u6570\u5b57\u8ba1\u7b97\u6982\u7387\u3002\u73b0\u5728\u5df2\u7ecf\u77e5\u9053\u4e86\u4e00\u4e2a\u8bcd\u662f\u5426\u51fa\u73b0\u5728\u4e00\u7bc7\u6587\u6863\u4e2d\uff0c\u4e5f\u77e5\u9053\u4e86\u8be5\u6587\u6863\u6240\u5c5e\u7684\u7c7b\u522b\u3002\u6240\u4ee5\uff0c\u6211\u4eec\u91cd\u5199\u4e4b\u524d\u7684\u8d1d\u53f6\u65af\u51c6\u5219\uff0c\u5c06\u4e4b\u524d\u7684x,y\u66ff\u6362\u6210w\uff0cw\u8868\u793a\u6587\u6863\u5411\u91cf\u3002\u8fd9\u4e2a\u4f8b\u5b50\u4e2d\uff0c\u6570\u503c\u4e2a\u5c5e\u4e8e\u8bcd\u6c47\u8868\u7684\u8bcd\u4e2a\u6570\u76f8\u540c\u3002<br \/>\n<a href=\"http:\/\/www.mrtblog.cn\/wp-content\/uploads\/2017\/02\/Bayes-formula-2.png\"><img loading=\"lazy\" decoding=\"async\" src=\"http:\/\/www.mrtblog.cn\/wp-content\/uploads\/2017\/02\/Bayes-formula-2.png\" alt=\"\" width=\"257\" height=\"69\" class=\"alignnone size-full wp-image-372\"><\/a><br \/>\n\u9996\u5148\uff0c\u901a\u8fc7\u7c7b\u522bi(\u4fae\u8fb1\u6027\u7559\u8a00\u6216\u6b63\u5e38\u7559\u8a00)\u4e2d\u6587\u6863\u6570\u9664\u4ee5\u603b\u7684\u6587\u6863\u6811\u6765\u8ba1\u7b97\u6982\u7387p(ci)\u3002\u63a5\u4e0b\u6765\u8ba1\u7b97p(w|ci)\uff0c\u8fd9\u5c31\u8981\u7528\u5230\u6734\u7d20\u8d1d\u53f6\u65af\u5047\u8bbe\u3002\u5982\u679c\u5c06w\u5c55\u5f00\u4e3a\u4e00\u4e2a\u4e2a\u72ec\u7acb\u7684\u7279\u5f81\uff0c\u90a3\u4e48\u5c31\u53ef\u4ee5\u5047\u8bbe\uff0c\u5b83\u610f\u5473\u7740\u53ef\u4ee5\u4f7f\u7528p(w0|ci)p(w1|ci)p(w2|ci)&#8230;p(wn|ci)\u6765\u8ba1\u7b97\u4e0a\u8ff0\u6982\u7387\u3002<\/p>\n<p>\u6253\u5f00Bayes.py\uff0c\u6dfb\u52a0\u5982\u4e0b\u4ee3\u7801\uff1a<\/p>\n<pre class=\"highlight\"><code class=\"language-python line-numbers\">\nfrom numpy import *\ndef trainNB0(trainMatrix,trainCategory):#\u8bad\u7ec3\u96c6\u6570\u7ec4\uff0c\u8bad\u7ec3\u96c6\u6807\u7b7e\n    numTrainDocs = len(trainMatrix)#\u8bad\u7ec3\u96c6\u603b\u957f\u5ea6\n    numWords = len(trainMatrix[0])#\u6bcf\u4e2a\u8bad\u7ec3\u96c6\u7684\u957f\u5ea6\n    pAbusive = sum(trainCategory) \/ float(numTrainDocs)#\u6807\u7b7e\u603b\u548c\/\u8bad\u7ec3\u6837\u672c\u6570\n    p0Num = zeros(numWords)#\u5c5e\u4e8e\u6b63\u5e38\u6587\u6863\n    p1Num = zeros(numWords)#\u5c5e\u4e8e\u4fae\u8fb1\u6027\u6587\u6863\n    p0Denom = 0.0\n    p1Denom = 0.0\n    for i in range(numTrainDocs):\n        if trainCategory[i] == 1:#\u4fae\u8fb1\u6027\u8bcd\u6c47\n            p1Num += trainMatrix[i]\n            p1Denom += sum(trainMatrix[i])#\u5f53\u524d\u7ec4\u6240\u6709\u6b21\u6570\u548c\n        else:#\u6b63\u5e38\u8bcd\u6c47\n            p0Num += trainMatrix[i]\n            p0Denom += sum(trainMatrix[i])\n    p1Vect = p1Num \/ p1Denom\n    p0Vect = p0Num \/ p0Denom\n    return p0Vect,p1Vect,pAbusive\n<\/code><\/pre>\n<p>\u4ee3\u7801\u8f93\u5165\u53c2\u6570\u4e3a\u6587\u6863\u77e9\u9635trainMatrix\uff0c\u4ee5\u53ca\u7531\u6bcf\u7bc7\u6587\u6863\u7c7b\u522b\u6807\u7b7e\u6240\u6784\u6210\u7684\u5411\u91cftrainCategory\u3002\u8ba1\u7b97\u4fae\u8fb1\u6027\u6587\u6863\u7684\u6982\u7387p(1)\uff0c\u518d\u8ba1\u7b97p(wi|c1)\u548cp(wi|c0)\u3002\u5b8c\u6210\u540e\u8f93\u5165\u6d4b\u8bd5\u4ee3\u7801\uff1a<\/p>\n<pre class=\"highlight\"><code class=\"language-python line-numbers\">\nimport Bayes\nlistOPosts,listClasses = Bayes.loadDataSet()\nmyVocabList = Bayes.createVocabList(listOPosts)\ntrainMat = []\nfor postinDoc in listOPosts:\n    trainMat.append(Bayes.setOfWords2Vec(myVocabList,postinDoc))\np0V,p1V,pAb = Bayes.trainNB0(trainMat,listClasses)\nprint(p0V)\nprint(p1V)\nprint(pAb)\n<\/code><\/pre>\n<p><a href=\"http:\/\/www.mrtblog.cn\/wp-content\/uploads\/2017\/02\/Bayes-example-2.png\"><img loading=\"lazy\" decoding=\"async\" src=\"http:\/\/www.mrtblog.cn\/wp-content\/uploads\/2017\/02\/Bayes-example-2.png\" alt=\"\" width=\"1228\" height=\"482\" class=\"alignnone size-full wp-image-373\" srcset=\"http:\/\/www.mrtblog.cn\/wp-content\/uploads\/2017\/02\/Bayes-example-2.png 1228w, http:\/\/www.mrtblog.cn\/wp-content\/uploads\/2017\/02\/Bayes-example-2-300x118.png 300w, http:\/\/www.mrtblog.cn\/wp-content\/uploads\/2017\/02\/Bayes-example-2-768x301.png 768w, http:\/\/www.mrtblog.cn\/wp-content\/uploads\/2017\/02\/Bayes-example-2-1024x402.png 1024w\" sizes=\"auto, (max-width: 1228px) 100vw, 1228px\" \/><\/a><br \/>\n\u53d1\u73b0\u6587\u6863\u5c5e\u4e8e\u4fae\u8fb1\u6027\u6587\u6863\u7684\u6982\u7387\u4e3a0.5.\u6240\u6709\u6982\u7387\u4e2d\u6700\u5927\u7684\u503c\u51fa\u73b0\u5728myVocabList\u7b2c26\u4e0b\u6807\u4f4d\u7f6e\uff0c\u5355\u8bcd\u4e3astupid\uff0c\u610f\u5473\u7740stupid\u6700\u80fd\u8868\u5f81\u7c7b\u522b1.<\/p>\n<h3>\u6d4b\u8bd5\u7b97\u6cd5\uff1a\u6839\u636e\u73b0\u5b9e\u60c5\u51b5\u4fee\u6539\u5206\u7c7b\u5668<\/h3>\n<p>\u5229\u7528\u8d1d\u53f6\u65af\u5206\u7c7b\u5668\u7684\u65f6\u5019\uff0c\u8981\u8ba1\u7b97\u591a\u4e2a\u6982\u7387\u7684\u6210\u7ee9\u4ee5\u83b7\u5f97\u6587\u6863\u8f93\u5165\u67d0\u4e2a\u7c7b\u522b\u7684\u6982\u7387\u3002\u5982\u679c\u5176\u4e2d\u4e00\u4e2a\u6982\u7387\u4e3a0\uff0c\u90a3\u4e48\u6700\u540e\u7684\u4e58\u79ef\u4e5f\u4e3a0.\u4e3a\u964d\u4f4e\u8fd9\u79cd\u5f71\u54cd\uff0c\u53ef\u4ee5\u5c06\u6240\u6709\u8bcd\u7684\u51fa\u73b0\u6570\u521d\u59cb\u5316\u4e3a1\uff0c\u5e76\u5c06\u5206\u6bcd\u521d\u59cb\u5316\u4e3a2.\u5c06trainNB0\u5bf9\u5e94\u4f4d\u7f6e\u6539\u4e3a\uff1a<\/p>\n<pre class=\"highlight\"><code class=\"language-python line-numbers\">\np0Num = ones(numWords)\np1Num = ones(numWords)\np0Denom = 2.0\np1Denom = 2.0\n<\/code><\/pre>\n<p>\u7531\u4e8e\u5927\u90e8\u5206\u56e0\u5b50\u90fd\u975e\u5e38\u5c0f\uff0c\u6240\u4ee5\u7a0b\u5e8f\u4f1a\u4e0b\u6ea2\u51fa\u6216\u8005\u5f97\u4e0d\u5230\u6b63\u786e\u7b54\u6848\uff0c\u89e3\u51b3\u529e\u6cd5\u4e3a\u5bf9\u4e58\u79ef\u53d6\u5bf9\u6570\uff0c\u5373\uff1a<\/p>\n<pre class=\"highlight\"><code class=\"language-python line-numbers\">\np1Vect = log(p1Num \/ p1Denom)\np0Vect = log(p0Num \/ p0Denom)\n<\/code><\/pre>\n<p>\u8fd9\u6837\u5c31\u6784\u5efa\u4e86\u5b8c\u6574\u7684\u5206\u7c7b\u5668\uff0c\u518d\u6253\u5f00Bayes.py\uff0c\u5c06\u4e0b\u5217\u4ee3\u7801\u6dfb\u52a0\u5230Bayes.py<\/p>\n<pre class=\"highlight\"><code class=\"language-python line-numbers\">\ndef classifyNB(vec2Classify,p0Vec,p1Vec,pClass1):\n    p1 = sum(vec2Classify * p1Vec) + log(pClass1)#\u5411\u91cf\u4e58\u79ef\u6c42\u548c\n    p0 = sum(vec2Classify * p0Vec) + log(1.0 - pClass1)\n    if (p1 &gt; p0):#\u8fd4\u56de\u5927\u6982\u7387\n        return 1\n    else:\n        return 0\n\ndef testingNB():#\u6d4b\u8bd5\u5206\u7c7b\u5668\n    listOPosts,listClasses = loadDataSet()\n    myVocabList = createVocabList(listOPosts)\n    trainMat = []\n    for postinDoc in listOPosts:\n        trainMat.append(setOfWords2Vec(myVocabList,postinDoc))\n        p0V,p1V,pAb = trainNB0(trainMat,listClasses)\n    testEntry=[&#039;love&#039;,&#039;my&#039;,&#039;dalmation&#039;]\n    thisDoc = array(setOfWords2Vec(myVocabList,testEntry))\n    print testEntry,&#039;classified as:&#039;,classifyNB(thisDoc,p0V,p1V,pAb)\n    testEntry = [&#039;stupid&#039;,&#039;garbage&#039;]\n    thisDoc = array(setOfWords2Vec(myVocabList,testEntry))\n    print testEntry,&#039;classified as: &#039;,classifyNB(thisDoc,p0V,p1V,pAb)\n<\/code><\/pre>\n<p>\u6d4b\u8bd5\u7ed3\u679c\uff1a<br \/>\n<a href=\"http:\/\/www.mrtblog.cn\/wp-content\/uploads\/2017\/02\/Bayes-example-3.png\"><img loading=\"lazy\" decoding=\"async\" src=\"http:\/\/www.mrtblog.cn\/wp-content\/uploads\/2017\/02\/Bayes-example-3.png\" alt=\"\" width=\"781\" height=\"161\" class=\"alignnone size-full wp-image-374\" srcset=\"http:\/\/www.mrtblog.cn\/wp-content\/uploads\/2017\/02\/Bayes-example-3.png 781w, http:\/\/www.mrtblog.cn\/wp-content\/uploads\/2017\/02\/Bayes-example-3-300x62.png 300w, http:\/\/www.mrtblog.cn\/wp-content\/uploads\/2017\/02\/Bayes-example-3-768x158.png 768w\" sizes=\"auto, (max-width: 781px) 100vw, 781px\" \/><\/a><\/p>\n","protected":false},"excerpt":{"rendered":"<p>1,345 Views\u53c2\u8003\u4e66\u7c4d\u300a\u673a\u5668\u5b66\u4e60\u5b9e\u6218\u300b KNN\u548c\u51b3\u7b56\u6811\u7b97\u6cd5\u8981\u6c42\u5206\u7c7b\u5668\u505a\u51fa\u8270\u96be\u51b3\u7b56\uff0c\u7ed9\u51fa\u201c\u8be5\u6570\u636e\u5b9e\u4f8b\u5c5e\u4e8e [&hellip;]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[105],"tags":[43,57],"class_list":["post-349","post","type-post","status-publish","format-standard","hentry","category-machine-learning","tag-python","tag-57"],"_links":{"self":[{"href":"http:\/\/www.mrtblog.cn\/index.php?rest_route=\/wp\/v2\/posts\/349","targetHints":{"allow":["GET"]}}],"collection":[{"href":"http:\/\/www.mrtblog.cn\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"http:\/\/www.mrtblog.cn\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"http:\/\/www.mrtblog.cn\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"http:\/\/www.mrtblog.cn\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=349"}],"version-history":[{"count":2,"href":"http:\/\/www.mrtblog.cn\/index.php?rest_route=\/wp\/v2\/posts\/349\/revisions"}],"predecessor-version":[{"id":353,"href":"http:\/\/www.mrtblog.cn\/index.php?rest_route=\/wp\/v2\/posts\/349\/revisions\/353"}],"wp:attachment":[{"href":"http:\/\/www.mrtblog.cn\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=349"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"http:\/\/www.mrtblog.cn\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=349"},{"taxonomy":"post_tag","embeddable":true,"href":"http:\/\/www.mrtblog.cn\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=349"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}