\u4e00\u53e5\u8bdd\u6765\u6982\u62ec\uff1a\u4eba\u4ee5\u7c7b\u805a\uff0c\u7269\u4ee5\u7fa4\u5206\u3002\u5c5e\u4e8e\u6709\u76d1\u7763\u5b66\u4e60<\/strong>\u3002<\/p>\n\n\n\n \u4e3e\u4e2a\u6817\u5b50\uff0c\u6211\u4eec\u53bb\u4e70\u6c34\u679c\uff0c\u4e3a\u4ec0\u4e48\u6211\u4eec\u80fd\u4ece\u4e00\u4e9b\u7279\u5f81\uff08Feature\uff09\u4e2d\u5206\u7c7b\u51fa\u6a58\u5b50\u548c\u6a59\u5b50\uff08\u6bd4\u5982\u91cd\u91cf\uff0c\u8f6f\u786c\uff0c\u8868\u76ae\u7684\u8936\u76b1\uff0c\u989c\u8272\u7684\u6df1\u6d45\uff09\uff1f\u56e0\u4e3a\u540c\u4e00\u7c7b\u6570\u636e\u4e4b\u95f4\u7684\u7279\u5f81\u66f4\u4e3a\u76f8\u4f3c\uff0c\u800c\u4e0d\u540c\u79cd\u7c7b\u6570\u636e\u4e4b\u95f4\u7684\u7279\u5f81\u5dee\u522b\u66f4\u5927\uff0c\u90a3\u4e48\u5982\u4f55\u7528\u6570\u5b66\u6216\u8005\u6a21\u578b\u6765\u5224\u65ad\u4ec0\u4e48\u53eb\u76f8\u4f3c\uff0c\u4ec0\u4e48\u53eb\u5dee\u522b\u66f4\u5927\uff1f<\/p>\n\n\n\n \u5728\u5206\u7c7b\u4efb\u52a1\u4e2d\uff0c\u6211\u4eec\u7684\u76ee\u6807\u662f\u5224\u65ad$x$\u7684\u7c7b\u522b$y$\u3002KNN\u4f1a\u5148\u89c2\u5bdf\u4e0e\u8be5\u6837\u672c\u70b9\u8ddd\u79bb\u6700\u8fd1\u7684$K$\u4e2a\u6837\u672c\uff0c\u7edf\u8ba1\u8fd9\u4e9b\u6837\u672c\u6240\u5c5e\u7684\u7c7b\u522b\uff0c\u5224\u65ad\u54ea\u4e00\u7c7b\u51fa\u73b0\u7684\u6b21\u6570\u6700\u591a\uff0c\u7136\u540e\u5c06\u8be5\u6837\u672c\u5f52\u7c7b\u5230\u6b21\u6570\u6700\u591a\u7684\u7c7b\u4e2d\u3002<\/p>\n\n\n\n \u4e3e\u4e2a\u6817\u5b50\uff1a\u5982\u56fe1 \u6240\u793a\uff0c\u5047\u8bbe\u73b0\u5728\u6709Class A \u548cClass B\u4ee5\u53ca\u8fd8\u672a\u5206\u7c7b\u7684\u7ea2\u8272\u4e94\u89d2\u661f\u3002\u76ee\u6807\u4efb\u52a1\u662f\u5224\u65ad\u5f53\u524d\u7ea2\u8272\u4e94\u89d2\u661f\u5c5e\u4e8e\u54ea\u4e2a\u7c7b\u522b\uff1f<\/p>\n\n\n\n \u6309\u7167\u601d\u8def\uff1a<\/p>\n\n\n\n \u4ece\u8fd9\u4e2a\u4f8b\u5b50\u53ef\u4ee5\u770b\u51faKNN\u7684\u57fa\u672c\u601d\u8def\u662f\u8ba9\u5f53\u524d\u6837\u672c\u7684\u5206\u7c7b\u670d\u4ece\u90bb\u5c45\u4e2d\u7684\u591a\u6570\u5206\u7c7b\u3002\u4f46\u662f\uff0c\u5f53K\u7684\u5927\u5c0f\u53d8\u5316\u65f6\uff0c\u7531\u4e8e\u90bb\u5c45\u6570\u91cf\u7684\u53d8\u5316\uff0c\u5176\u591a\u5c11\u7c7b\u522b\u4e5f\u4f1a\u8ddf\u7740\u53d1\u751f\u53d8\u5316\uff0c\u4ece\u800c\u5f71\u54cd\u5230\u5bf9\u5f53\u524d\u6837\u672c\u7684\u5206\u7c7b\u5224\u65ad\u3002<\/p>\n\n\n\n \u56e0\u6b64\uff0c\u51b3\u5b9aK\u7684\u5927\u5c0f\u662fKNN\u6700\u91cd\u8981\u7684\u90e8\u5206\u4e4b\u4e00\u3002\u5f53K\u7684\u53d6\u503c\u592a\u5c0f\uff0c\u5206\u7c7b\u7ed3\u679c\u5f88\u5bb9\u6613\u53d7\u5230\u6837\u672c\u5468\u56f4\u4e2a\u522b\u566a\u58f0\u6570\u636e\u7684\u5f71\u54cd\uff0c\u5bb9\u6613\u53d1\u751f\u8fc7\u62df\u5408\uff1b\u5f53K\u53d6\u7684\u592a\u5927\uff0c\u53c8\u53ef\u80fd\u5c06\u8fdc\u5904\u4e00\u4e9b\u4e0d\u76f8\u5e72\u7684\u6837\u672c\u6570\u636e\u5305\u542b\u8fdb\u6765\uff0c\u5bb9\u6613\u53d1\u751f\u6b20\u62df\u5408\uff0c\u90a3\u4e48\u5982\u4f55\u6765\u786e\u5b9aK\u503c\u5462\uff1f\u4ee5\u53ca\u5982\u4f55\u5ea6\u91cf\u8fd9\u4e2a\u8ddd\u79bb\uff1f<\/p>\n\n\n\n \u91cd\u8981\u53c2\u6570\u7684\u786e\u5b9a\uff1a<\/p>\n\n\n\n \u786e\u5b9aK\u503c\u7684\u65b9\u6cd5\u6709\uff1a<\/p>\n\n\n\n \u539f\u7406<\/strong>\uff1a\u5c06\u6570\u636e\u96c6\u5212\u5206\u4e3a\u591a\u4e2a\u5b50\u96c6\uff0c\u5728\u6bcf\u6b21\u9a8c\u8bc1\u8fc7\u7a0b\u4e2d\uff0c\u9009\u4e00\u4e2a\u5b50\u96c6\u4f5c\u4e3a\u9a8c\u8bc1\u96c6\uff0c\u5176\u4f59\u5b50\u96c6\u4f5c\u4e3a\u8bad\u7ec3\u96c6\uff0c\u5bf9\u4e8e\u4e0d\u540c\u7684K\u503c\u8fdb\u884c\u6a21\u578b\u8bad\u7ec3\u548c\u8bc4\u4f30\uff0c\u6700\u540e\u8ba1\u7b97\u6bcf\u4e2aK\u503c\u5728\u6240\u6709\u9a8c\u8bc1\u96c6\u4e0a\u7684\u5e73\u5747\u6027\u80fd\uff0c\u9009\u6027\u80fd\u6700\u4f18\u7684K\u503c\uff08\u5c31\u662f\u5c1d\u8bd5\u591a\u4e2a\u4e0d\u540c\u7684K\u503c<\/strong>\uff0c\u5e76\u901a\u8fc7\u5728\u9a8c\u8bc1\u96c6\u4e0a<\/strong>\u7684\u8868\u73b0\uff08\u5982\u51c6\u786e\u7387\u3001F1\u5206\u6570\u7b49\uff09\u6765\u9009\u62e9\u8868\u73b0\u6700\u597d\u7684 K \u503c<\/strong>\uff09\u3002<\/p>\n\n\n\n \u6848\u4f8b<\/strong>\uff1a\uff08\u9e22\u5c3e\u82b1\u6570\u636e\u96c6YYDS\uff09<\/p>\n\n\n\n \u539f\u7406<\/strong>\uff1a\u8be5\u65b9\u6cd5\u4f1a\u5728\u4e00\u4e2a\u9884\u5148\u8bbe\u5b9a\u597d\u7684 K \u503c\u8303\u56f4\u5185\uff0c\u5bf9\u6bcf\u4e2a\u53ef\u80fd\u7684 K \u503c\u8fdb\u884c\u5c1d\u8bd5\uff0c\u4f7f\u7528\u4ea4\u53c9\u9a8c\u8bc1\u6765\u8bc4\u4f30\u6a21\u578b\u5728\u4e0d\u540c K \u503c\u4e0b\u7684\u6027\u80fd\uff0c\u6700\u7ec8\u9009\u53d6\u6027\u80fd\u6700\u4f18\u7684 K \u503c\u3002<\/p>\n\n\n\n \u6848\u4f8b<\/strong>\uff1a<\/p>\n\n\n\n \u5728 KNN\uff08K – \u6700\u8fd1\u90bb\uff09\u7b97\u6cd5\u4e2d\uff0c\u8ddd\u79bb\u7684\u786e\u5b9a\u81f3\u5173\u91cd\u8981\uff0c\u5b83\u7528\u4e8e\u8861\u91cf\u6570\u636e\u70b9\u4e4b\u95f4\u7684\u76f8\u4f3c\u6027\uff0c\u8fdb\u800c\u627e\u51fa\u5f85\u9884\u6d4b\u70b9\u7684 K \u4e2a\u6700\u8fd1\u90bb\u3002\u4ee5\u4e0b\u662f\u51e0\u79cd\u5e38\u89c1\u7684\u8ddd\u79bb\u5ea6\u91cf\u7b97\u6cd5\uff1a<\/p>\n\n\n\n \u5b9a\u4e49<\/strong>\uff1a\u6b27\u6c0f\u8ddd\u79bb\u662f\u6700\u5e38\u7528\u7684\u8ddd\u79bb\u5ea6\u91cf\u65b9\u6cd5\u4e4b\u4e00\uff0c\u5b83\u57fa\u4e8e\u52fe\u80a1\u5b9a\u7406\uff0c\u7528\u4e8e\u8ba1\u7b97 n \u7ef4\u7a7a\u95f4\u4e2d\u4e24\u4e2a\u70b9\u4e4b\u95f4\u7684\u76f4\u7ebf\u8ddd\u79bb\u3002<\/p>\n\n\n\n \u516c\u5f0f<\/strong>\uff1a\u5bf9\u4e8e\u4e24\u4e2a n \u7ef4\u5411\u91cf$\\vec{x} = (x_1,x_2,\\cdots,x_n)$\u548c$\\vec{y} = (y_1,y_2,\\cdots,y_n)$\uff0c\u90a3\u4ed6\u4eec\u4e4b\u95f4\u7684\u6b27\u5f0f\u8ddd\u79bb$d(\\vec{x},\\vec{y})$ \u4e3a\uff1a<\/p>\n\n\n\n $$d(\\vec{x},\\vec{y}) = \\sqrt{\\sum_{i = 1}^{n}(x_i – y_i)^2}$$<\/p>\n\n\n\n \u6848\u4f8b<\/strong>\uff1a<\/p>\n\n\n\n \u5e94\u7528\u573a\u666f<\/strong>\uff1a\u9002\u7528\u4e8e\u6570\u636e\u5206\u5e03\u8f83\u4e3a\u5747\u5300\u3001\u7279\u5f81\u4e4b\u95f4\u76f8\u4e92\u72ec\u7acb\u7684\u60c5\u51b5\uff0c\u5728\u5f88\u591a\u5b9e\u9645\u95ee\u9898\u4e2d\u90fd\u6709\u5e7f\u6cdb\u5e94\u7528\uff0c\u5982\u5730\u7406\u5750\u6807\u8ddd\u79bb\u8ba1\u7b97\u3001\u56fe\u50cf\u50cf\u7d20\u76f8\u4f3c\u5ea6\u8ba1\u7b97\u7b49\u3002<\/p>\n\n\n\n \u5b9a\u4e49<\/strong>\uff1a\u4e5f\u79f0\u4e3a\u57ce\u5e02\u8857\u533a\u8ddd\u79bb\uff0c\u5b83\u662f\u6307\u5728\u7f51\u683c\u72b6\u7684\u57ce\u5e02\u8857\u9053\u4e2d\uff0c\u4ece\u4e00\u4e2a\u70b9\u5230\u53e6\u4e00\u4e2a\u70b9\u6cbf\u7740\u8857\u9053\u884c\u8d70\u7684\u6700\u77ed\u8ddd\u79bb\uff08\u7c7b\u4f3c\u4e0b\u56f4\u68cb\uff0c\u53ea\u80fdx\u548cy\u65b9\u5411\u8d70\u7f51\u683c\u70b9\uff09\u3002<\/p>\n\n\n\n \u516c\u5f0f<\/strong>\uff1a\u5bf9\u4e8e\u4e24\u4e2a n \u7ef4\u5411\u91cf$\\vec{x} = (x_1,x_2,\\cdots,x_n)$\u548c$\\vec{y} = (y_1,y_2,\\cdots,y_n)$\uff0c\u90a3\u4ed6\u4eec\u4e4b\u95f4\u7684\u66fc\u54c8\u987f\u8ddd\u79bb$d(\\vec{x},\\vec{y})$ \u4e3a\uff1a<\/p>\n\n\n\n $$d(\\vec{x},\\vec{y}) = \\sum_{i = 1}^{n}|x_i – y_i|$$<\/p>\n\n\n\n \u6848\u4f8b<\/strong>\uff1a<\/p>\n\n\n\n \u5b9a\u4e49<\/strong>\uff1a\u95f5\u53ef\u592b\u65af\u57fa\u8ddd\u79bb\u662f\u6b27\u6c0f\u8ddd\u79bb\u548c\u66fc\u54c8\u987f\u8ddd\u79bb\u7684\u63a8\u5e7f\uff0c\u5b83\u5f15\u5165\u4e86\u4e00\u4e2a\u53c2\u6570 p\uff0c\u901a\u8fc7\u6539\u53d8 p \u7684\u503c\u53ef\u4ee5\u5f97\u5230\u4e0d\u540c\u7684\u8ddd\u79bb\u5ea6\u91cf\u3002<\/p>\n\n\n\n \u516c\u5f0f<\/strong>\uff1a\u5bf9\u4e8e\u4e24\u4e2a n \u7ef4\u5411\u91cf$\\vec{x} = (x_1,x_2,\\cdots,x_n)$\u548c$\\vec{y} = (y_1,y_2,\\cdots,y_n)$\uff0c\u4ed6\u4eec\u4e4b\u95f4\u7684\u95f5\u53ef\u592b\u65af\u57fa\u8ddd\u79bb$d(\\vec{x},\\vec{y})$ \u4e3a\uff1a<\/p>\n\n\n\n $$d(\\vec{x},\\vec{y}) = (\\sum_{i = 1}^{n}(x_i – y_i)^p)^{1\/p}$$<\/p>\n\n\n\n \u5176\u4e2d\uff0c$p>=1$\u3002\u5f53$p = 2$\u65f6\uff0c\u95f5\u53ef\u592b\u65af\u57fa\u8ddd\u79bb\u5c31\u662f\u6b27\u6c0f\u8ddd\u79bb\uff1b\u5f53$p = 1$\u65f6\uff0c\u5c31\u662f\u66fc\u54c8\u987f\u8ddd\u79bb\u3002<\/p>\n\n\n\n \u6848\u4f8b<\/strong>\uff1a<\/p>\n\n\n\n \u5b9a\u4e49<\/strong>\uff1a\u5207\u6bd4\u96ea\u592b\u8ddd\u79bb\u662f\u6307\u5728 n \u7ef4\u7a7a\u95f4\u4e2d\uff0c\u4e24\u4e2a\u70b9\u5728\u5404\u4e2a\u5750\u6807\u4e0a\u7684\u5dee\u503c\u7684\u6700\u5927\u503c\u3002<\/p>\n\n\n\n \u516c\u5f0f<\/strong>\uff1a\u5bf9\u4e8e\u4e24\u4e2a n \u7ef4\u5411\u91cf$\\vec{x} = (x_1,x_2,\\cdots,x_n)$\u548c$\\vec{y} = (y_1,y_2,\\cdots,y_n)$\uff0c\u4ed6\u4eec\u4e4b\u95f4\u7684\u5207\u6bd4\u96ea\u592b\u8ddd\u79bb$d(\\vec{x},\\vec{y})$ \u4e3a\uff1a<\/p>\n\n\n\n $$d(\\vec{x},\\vec{y}) = \\max_{i = 1}^{n}|x_i – y_i|$$<\/p>\n\n\n\n \u6848\u4f8b<\/strong>\uff1a<\/p>\n\n\n\n \u5b9a\u4e49<\/strong>\uff1a\u4f59\u5f26\u76f8\u4f3c\u5ea6\u901a\u8fc7\u8ba1\u7b97\u4e24\u4e2a\u5411\u91cf\u7684\u5939\u89d2\u4f59\u5f26\u503c\u6765\u8861\u91cf\u5b83\u4eec\u7684\u65b9\u5411\u76f8\u4f3c\u5ea6\uff0c\u503c\u8d8a\u63a5\u8fd1 1 \u8868\u793a\u65b9\u5411\u8d8a\u76f8\u4f3c\u3002\u4f59\u5f26\u8ddd\u79bb\u5219\u662f\u7528 1 \u51cf\u53bb\u4f59\u5f26\u76f8\u4f3c\u5ea6\u5f97\u5230\uff0c\u7528\u4e8e\u8868\u793a\u5411\u91cf\u4e4b\u95f4\u7684\u5dee\u5f02\u7a0b\u5ea6\u3002<\/p>\n\n\n\n \u516c\u5f0f<\/strong>\uff1a\u5bf9\u4e8e\u4e24\u4e2a n \u7ef4\u5411\u91cf$\\vec{x} = (x_1,x_2,\\cdots,x_n)$\u548c$\\vec{y} = (y_1,y_2,\\cdots,y_n)$\uff0c\u4ed6\u4eec\u4e4b\u95f4\u7684\u4f59\u5f26\u76f8\u4f3c\u5ea6\u8ddd\u79bb$sim(\\vec{x},\\vec{y})$ \u4e3a\uff1a<\/p>\n\n\n\n $$sim(\\vec{x},\\vec{y}) = \\frac{\\mathbf{A}\\cdot\\mathbf{B}}{\\|\\mathbf{A}\\|\\|\\mathbf{B}\\|}=\\frac{\\sum_{i = 1}^{n}A_iB_i}{\\sqrt{\\sum_{i = 1}^{n}A_{i}^{2}}\\sqrt{\\sum_{i = 1}^{n}B_{i}^{2}}}$$<\/p>\n\n\n\n \u4f59\u5f26\u8ddd\u79bb\u00a0: $d(\\vec{x},\\vec{y}) = 1 – sim(\\vec{x},\\vec{y})$<\/p>\n\n\n\n \u6848\u4f8b<\/strong>\uff1a<\/p>\n\n\n\n \u5e94\u7528\u573a\u666f<\/strong>\uff1a\u5e38\u7528\u4e8e\u6587\u672c\u6316\u6398\u3001\u4fe1\u606f\u68c0\u7d22\u7b49\u9886\u57df\uff0c\u56e0\u4e3a\u5b83\u66f4\u5173\u6ce8\u5411\u91cf\u7684\u65b9\u5411\u800c\u4e0d\u662f\u5411\u91cf\u7684\u957f\u5ea6\uff0c\u9002\u5408\u5904\u7406\u9ad8\u7ef4\u7a00\u758f\u6570\u636e\u3002<\/p>\n\n\n\n 1.1 \u533b\u7597\u8bca\u65ad<\/strong>\uff1a<\/p>\n\n\n\n \u539f\u7406<\/strong>\uff1a\u6839\u636e\u60a3\u8005\u7684\u75c7\u72b6\u3001\u68c0\u67e5\u7ed3\u679c\u7b49\u7279\u5f81\u6784\u5efa\u7279\u5f81\u5411\u91cf\u3002\u5bf9\u4e8e\u4e00\u4e2a\u65b0\u60a3\u8005\uff0c\u627e\u5230\u4e0e\u4ed6\u7279\u5f81\u6700\u76f8\u4f3c\u7684 K \u4e2a\u60a3\u8005\uff0c\u6839\u636e\u8fd9 K \u4e2a\u60a3\u8005\u7684\u8bca\u65ad\u7ed3\u679c\u6765\u63a8\u65ad\u65b0\u60a3\u8005\u53ef\u80fd\u60a3\u6709\u7684\u75be\u75c5\u3002<\/p>\n\n\n\n \u793a\u4f8b<\/strong>\uff1a\u6839\u636e\u60a3\u8005\u7684\u4f53\u6e29\u3001\u8840\u538b\u3001\u8840\u6db2\u6307\u6807\u7b49\u4fe1\u606f\uff0c\u4f7f\u7528 KNN \u7b97\u6cd5\u8f85\u52a9\u533b\u751f\u5224\u65ad\u60a3\u8005\u662f\u5426\u60a3\u6709\u67d0\u79cd\u75be\u75c5\uff0c\u5982\u7cd6\u5c3f\u75c5\u3001\u5fc3\u810f\u75c5\u7b49\u3002<\/p>\n\n\n\n 1.2 \u5546\u54c1\u63a8\u8350<\/strong><\/p>\n\n\n\n \u539f\u7406<\/strong>\uff1a\u6839\u636e\u7528\u6237\u7684\u5386\u53f2\u8d2d\u4e70\u8bb0\u5f55\u3001\u6d4f\u89c8\u884c\u4e3a\u3001\u8bc4\u4ef7\u4fe1\u606f\u7b49\u6784\u5efa\u7528\u6237\u7684\u7279\u5f81\u5411\u91cf\u3002\u5bf9\u4e8e\u4e00\u4e2a\u7528\u6237\uff0c\u627e\u5230\u4e0e\u4ed6\u5174\u8da3\u6700\u76f8\u4f3c\u7684 K \u4e2a\u7528\u6237\uff0c\u6839\u636e\u8fd9 K \u4e2a\u7528\u6237\u8d2d\u4e70\u8fc7\u7684\u5546\u54c1\u6765\u4e3a\u8be5\u7528\u6237\u63a8\u8350\u5546\u54c1\u3002<\/p>\n\n\n\n \u793a\u4f8b<\/strong>\uff1a\u7535\u5546\u5e73\u53f0\u6839\u636e\u7528\u6237\u7684\u8d2d\u7269\u5386\u53f2\u548c\u504f\u597d\uff0c\u4f7f\u7528 KNN \u7b97\u6cd5\u4e3a\u7528\u6237\u63a8\u8350\u53ef\u80fd\u611f\u5174\u8da3\u7684\u5546\u54c1\uff0c\u63d0\u9ad8\u7528\u6237\u7684\u8d2d\u4e70\u8f6c\u5316\u7387\u548c\u6ee1\u610f\u5ea6\u3002<\/p>\n\n\n\n 2.1 \u623f\u4ef7\u9884\u6d4b<\/strong><\/p>\n\n\n\n \u539f\u7406<\/strong>\uff1a\u5c06\u623f\u5c4b\u7684\u5404\u79cd\u7279\u5f81\uff08\u5982\u9762\u79ef\u3001\u623f\u95f4\u6570\u91cf\u3001\u623f\u9f84\u3001\u5730\u7406\u4f4d\u7f6e\u7b49\uff09\u4f5c\u4e3a\u7279\u5f81\u5411\u91cf\u3002\u5bf9\u4e8e\u4e00\u4e2a\u5f85\u9884\u6d4b\u623f\u4ef7\u7684\u623f\u5c4b\uff0c\u627e\u5230\u4e0e\u5b83\u7279\u5f81\u6700\u76f8\u4f3c\u7684 K \u4e2a\u623f\u5c4b\uff0c\u8ba1\u7b97\u8fd9 K \u4e2a\u623f\u5c4b\u7684\u5e73\u5747\u623f\u4ef7\u6216\u52a0\u6743\u5e73\u5747\u623f\u4ef7\uff0c\u5c06\u5176\u4f5c\u4e3a\u5f85\u9884\u6d4b\u623f\u5c4b\u7684\u623f\u4ef7\u3002<\/p>\n\n\n\n \u793a\u4f8b<\/strong>\uff1a\u5728\u623f\u5730\u4ea7\u5e02\u573a\u4e2d\uff0c\u901a\u8fc7\u6536\u96c6\u5927\u91cf\u623f\u5c4b\u7684\u76f8\u5173\u4fe1\u606f\u548c\u5b9e\u9645\u6210\u4ea4\u4ef7\u683c\uff0c\u4f7f\u7528 KNN \u7b97\u6cd5\u53ef\u4ee5\u4e3a\u65b0\u4e0a\u5e02\u7684\u623f\u5c4b\u63d0\u4f9b\u5408\u7406\u7684\u4ef7\u683c\u9884\u6d4b\u3002<\/p>\n\n\n\n 2.2 \u80a1\u7968\u4ef7\u683c\u9884\u6d4b<\/strong><\/p>\n\n\n\n \u539f\u7406<\/strong>\uff1a\u9009\u53d6\u80a1\u7968\u7684\u5386\u53f2\u4ef7\u683c\u3001\u6210\u4ea4\u91cf\u3001\u5e02\u76c8\u7387\u7b49\u7279\u5f81\u6784\u5efa\u7279\u5f81\u5411\u91cf\u3002\u5bf9\u4e8e\u672a\u6765\u67d0\u4e00\u65f6\u523b\u7684\u80a1\u7968\u4ef7\u683c\u9884\u6d4b\uff0c\u627e\u5230\u4e0e\u5f53\u524d\u65f6\u523b\u7279\u5f81\u6700\u76f8\u4f3c\u7684 K \u4e2a\u5386\u53f2\u65f6\u523b\uff0c\u6839\u636e\u8fd9 K \u4e2a\u65f6\u523b\u4e4b\u540e\u7684\u80a1\u7968\u4ef7\u683c\u53d8\u5316\u60c5\u51b5\u6765\u9884\u6d4b\u672a\u6765\u7684\u80a1\u7968\u4ef7\u683c\u3002<\/p>\n\n\n\n \u793a\u4f8b<\/strong>\uff1a\u6295\u8d44\u8005\u53ef\u4ee5\u5229\u7528 KNN \u7b97\u6cd5\u7ed3\u5408\u80a1\u7968\u7684\u5386\u53f2\u6570\u636e\uff0c\u5bf9\u80a1\u7968\u7684\u672a\u6765\u8d70\u52bf\u8fdb\u884c\u9884\u6d4b\uff0c\u8f85\u52a9\u6295\u8d44\u51b3\u7b56\u3002<\/p>\n\n\n\n \u9e22\u5c3e\u82b1\u6570\u636e\u96c6\u662f\u4e00\u4e2a\u975e\u5e38\u7ecf\u5178\u4e14\u5e38\u7528\u7684\u6570\u636e\u96c6\uff0c\u5305\u62ec150\u4e2a\u6837\u672c\uff0c\u5c5e\u4e8e\u591a\u5206\u7c7b\u95ee\u9898\u7684\u6570\u636e\u96c6\u3002<\/p>\n\n\n\n \u7279\u5f81<\/strong>\uff1a\u8be5\u6570\u636e\u96c6\u4e3b\u8981\u6709 4 \u4e2a\u6570\u503c\u578b\u7279\u5f81\uff0c\u7528\u4e8e\u63cf\u8ff0\u9e22\u5c3e\u82b1\u7684\u7269\u7406\u7279\u5f81\u3002<\/p>\n\n\n\n \u6807\u7b7e<\/strong>\uff1a\u6837\u672c\u88ab\u5206\u4e3a 3 \u4e2a\u4e0d\u540c\u7684\u7c7b\u522b\uff0c\u4ee3\u8868 3 \u79cd\u4e0d\u540c\u79cd\u7c7b\u7684\u9e22\u5c3e\u82b1\u3002<\/p>\n\n\n\n \u6848\u4f8b<\/strong>\uff1a\u76f4\u63a5\u7528sklearn \u5bfc\u5165iris\u6570\u636e\u96c6<\/p>\n\n\n\n \u6570\u636e\u7ed3\u679c<\/strong>\uff1a<\/p>\n\n\n \u8981\u6c42<\/strong>\uff1a\u4f7f\u7528iris\u6570\u636e\u96c6\uff0c\u7ed3\u5408KNN\u7b97\u6cd5\uff08\u4f7f\u7528\u66fc\u54c8\u987f\u8ddd\u79bb\u7b97\u6cd5\uff09\u8fdb\u884c\u9e22\u5c3e\u82b1\u7684\u5206\u7c7b\uff1b\u751f\u6210\u597d\u6a21\u578b\u4e4b\u540e\uff0c\u968f\u673a\u7ed9\u51e0\u7ec4\u7279\u5f81\u5411\u91cf\uff0c\u6765\u68c0\u67e5\u5206\u7c7b\u60c5\u51b5\u3002<\/p>\n\n\n\n \u7ed3\u679c\u4e3a\uff1a<\/p>\n\n\n\n <\/p>\n","protected":false},"excerpt":{"rendered":" \u4e00\u3001\u524d\u8a00 \u673a\u5668\u5b66\u4e60\u5e76\u4e0d\u5168\u662f\u795e\u7ecf\u7f51\u7edc\uff0c\u76f8\u5f53\u591a\u7684\u673a\u5668\u5b66\u4e60\u7b97\u6cd5\u90fd\u975e\u5e38\u7b80\u5355\u548c\u76f4\u89c2\uff0c\u4e00\u4e2a\u4e00\u4e2a\u6162\u6162\u5b66\u3002\u5148\u5b66\u4e60\u4e00\u4e2a\u6700\u57fa\u672c\u7684\u5206 […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[26],"tags":[32],"class_list":["post-245","post","type-post","status-publish","format-standard","hentry","category-ml","tag-32"],"_links":{"self":[{"href":"https:\/\/zhangliangshan.cn\/index.php\/wp-json\/wp\/v2\/posts\/245","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/zhangliangshan.cn\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/zhangliangshan.cn\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/zhangliangshan.cn\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/zhangliangshan.cn\/index.php\/wp-json\/wp\/v2\/comments?post=245"}],"version-history":[{"count":25,"href":"https:\/\/zhangliangshan.cn\/index.php\/wp-json\/wp\/v2\/posts\/245\/revisions"}],"predecessor-version":[{"id":289,"href":"https:\/\/zhangliangshan.cn\/index.php\/wp-json\/wp\/v2\/posts\/245\/revisions\/289"}],"wp:attachment":[{"href":"https:\/\/zhangliangshan.cn\/index.php\/wp-json\/wp\/v2\/media?parent=245"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/zhangliangshan.cn\/index.php\/wp-json\/wp\/v2\/categories?post=245"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/zhangliangshan.cn\/index.php\/wp-json\/wp\/v2\/tags?post=245"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}
2. \u56de\u5f52\uff08Regression\uff09\uff1a\u662f\u6307\u9884\u6d4b\u4e00\u4e2a\u8fde\u7eed\u6570\u503c\u7684\u4efb\u52a1\u3002\u6a21\u578b\u901a\u8fc7\u5b66\u4e60\u8bad\u7ec3\u6570\u636e\u4e2d\u7684\u7279\u5f81\u548c\u5bf9\u5e94\u7684\u8fde\u7eed\u76ee\u6807\u503c\u4e4b\u95f4\u7684\u5173\u7cfb\uff0c\u6765\u5bf9\u65b0\u7684\u6570\u636e\u6837\u672c\u7684\u76ee\u6807\u503c\u8fdb\u884c\u9884\u6d4b\u3002\u7b80\u5355\u6765\u8bf4\uff0c\u5c31\u662f\u9884\u6d4b\u4e00\u4e2a\u5177\u4f53\u7684\u6570\u503c<\/strong>\u3002
3. \u7279\u5f81\uff08Feature\uff09\uff1a\u662f\u63cf\u8ff0\u6570\u636e\u6837\u672c\u7684\u5404\u79cd\u5c5e\u6027<\/strong>\u6216\u53d8\u91cf<\/strong>\uff0c\u5b83\u662f\u6a21\u578b\u8fdb\u884c\u5b66\u4e60\u548c\u9884\u6d4b\u7684\u57fa\u7840\u3002
4. \u6709\u76d1\u7763\u5b66\u4e60\uff08supervised Learning\uff09\uff1a\u662f\u6307\u5728\u8bad\u7ec3\u6a21\u578b\u65f6\uff0c\u4f7f\u7528\u5305\u542b\u7279\u5f81\u548c\u5bf9\u5e94\u6807\u7b7e\uff08\u76ee\u6807\u503c\uff09\u7684\u6570\u636e\u8fdb\u884c\u5b66\u4e60\u3002\u6a21\u578b\u7684\u76ee\u6807\u662f\u901a\u8fc7\u5b66\u4e60\u7279\u5f81\u4e0e\u6807\u7b7e\u4e4b\u95f4\u7684\u6620\u5c04\u5173\u7cfb\uff0c\u5bf9\u65b0\u7684\u672a\u77e5\u6570\u636e\u8fdb\u884c\u51c6\u786e\u7684\u9884\u6d4b\u3002\u7b80\u5355\u6765\u8bf4\uff0c\u5c31\u662f\u7ed9\u6a21\u578b\u63d0\u4f9b \u201c\u6807\u51c6\u7b54\u6848\u201d \u8fdb\u884c\u5b66\u4e60\u3002
5. \u65e0\u76d1\u7763\u5b66\u4e60\uff08Unsupervised Learning\uff09\uff1a\u662f\u6307\u5728\u8bad\u7ec3\u6a21\u578b\u65f6\uff0c\u4f7f\u7528\u4ec5\u5305\u542b\u7279\u5f81\u800c\u6ca1\u6709\u5bf9\u5e94\u6807\u7b7e\u7684\u6570\u636e\u8fdb\u884c\u5b66\u4e60\u3002\u6a21\u578b\u7684\u76ee\u6807\u662f\u4ece\u6570\u636e\u4e2d\u53d1\u73b0\u6f5c\u5728\u7684\u7ed3\u6784\u3001\u6a21\u5f0f\u6216\u89c4\u5f8b\u3002\u7531\u4e8e\u6ca1\u6709 \u201c\u6807\u51c6\u7b54\u6848\u201d\uff0c\u6a21\u578b\u9700\u8981\u81ea\u4e3b\u5730\u5bf9\u6570\u636e\u8fdb\u884c\u5206\u6790\u548c\u7406\u89e3\u3002<\/span><\/div>\n\n\n\n\u4e8c\u3001KNN\u7b97\u6cd5\u7684\u539f\u7406<\/h2>\n\n\n\n
\n
![\"\"](\"https:\/\/zhangliangshan.cn\/wp-content\/uploads\/2025\/02\/image-25-1024x760.png\")
<\/div>\n
\u4e09\u3001\u9009\u62e9K\u503c<\/h2>\n\n\n\n
1. \u4ea4\u53c9\u9a8c\u8bc1\u6cd5\uff1a<\/h3>\n\n\n\n
<\/ol>\n\n\n\n
from sklearn.neighbors import KNeighborsClassifier\nfrom sklearn.model_selection import cross_val_score\nfrom sklearn.datasets import load_iris\nimport numpy as np\n\n# \u52a0\u8f7d\u6570\u636e\u96c6\niris = load_iris()\nX = iris.data\ny = iris.target\n\n# \u5b9a\u4e49K\u503c\u7684\u8303\u56f4\nk_range = range(1, 31)\nk_scores = []\n\n# \u5bf9\u6bcf\u4e2aK\u503c\u8fdb\u884c\u4ea4\u53c9\u9a8c\u8bc1\nfor k in k_range:\n knn = KNeighborsClassifier(n_neighbors=k)\n scores = cross_val_score(knn, X, y, cv=5, scoring='accuracy')\n k_scores.append(scores.mean())\n\n# \u627e\u5230\u6700\u4f73K\u503c\nbest_k = np.argmax(k_scores) + 1\nprint(f\"\u6700\u4f73K\u503c\u4e3a: {best_k}\")<\/code><\/pre>\n\n\n\n2. \u7f51\u683c\u641c\u7d22\u6cd5<\/h3>\n\n\n\n
from sklearn.neighbors import KNeighborsClassifier\nfrom sklearn.model_selection import GridSearchCV\nfrom sklearn.datasets import load_iris\nfrom sklearn.model_selection import train_test_split\n\n# \u52a0\u8f7d\u6570\u636e\u96c6\niris = load_iris()\nX = iris.data\ny = iris.target\n\n# \u5212\u5206\u8bad\u7ec3\u96c6\u548c\u6d4b\u8bd5\u96c6\nX_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=42)\n\n# \u5b9a\u4e49K\u503c\u7684\u641c\u7d22\u8303\u56f4\nparam_grid = {'n_neighbors': range(1, 21)}\n\n# \u521b\u5efaKNN\u5206\u7c7b\u5668\nknn = KNeighborsClassifier()\n\n# \u4f7f\u7528\u7f51\u683c\u641c\u7d22\u8fdb\u884c\u53c2\u6570\u8c03\u4f18\ngrid_search = GridSearchCV(knn, param_grid, cv=5)\ngrid_search.fit(X_train, y_train)\n\n# \u8f93\u51fa\u6700\u4f73K\u503c\nbest_k = grid_search.best_params_['n_neighbors']\nprint(f\"\u6700\u4f73K\u503c\u4e3a: {best_k}\")<\/code><\/pre>\n\n\n\n3. \u7ecf\u9a8c\u6cd5\uff0c\u52a0\u6743\u5e73\u5747\u7b49\u7b49<\/h3>\n\n\n\n
\u56db\u3001\u5ea6\u91cf\u8ddd\u79bb<\/h2>\n\n\n\n
1. \u6b27\u5f0f\u8ddd\u79bb\uff08Euclidean Distance\uff09<\/h3>\n\n\n\n
import numpy as np\n\nx = np.array([1, 2, 3])\ny = np.array([4, 5, 6])\n\ndistance = np.sqrt(np.sum((x - y) ** 2))\nprint(f\"\u6b27\u6c0f\u8ddd\u79bb: {distance}\")<\/code><\/pre>\n\n\n\n2. \u66fc\u54c8\u987f\u8ddd\u79bb\uff08Manhattan Distance\uff09<\/h3>\n\n\n\n
import numpy as np\n\nx = np.array([1, 2, 3])\ny = np.array([4, 5, 6])\n\ndistance = np.sum(np.abs(x - y))\nprint(f\"\u66fc\u54c8\u987f\u8ddd\u79bb: {distance}\")<\/code><\/pre>\n\n\n\n3. \u95f5\u53ef\u592b\u65af\u57fa\u8ddd\u79bb\uff08Minkowski Distance\uff09<\/h3>\n\n\n\n
import numpy as np\nfrom scipy.spatial.distance import minkowski\n\nx = np.array([1, 2, 3])\ny = np.array([4, 5, 6])\np = 3\n\ndistance = minkowski(x, y, p)\nprint(f\"\u95f5\u53ef\u592b\u65af\u57fa\u8ddd\u79bb (p={p}): {distance}\")<\/code><\/pre>\n\n\n\n4. \u5207\u6bd4\u96ea\u592b\u8ddd\u79bb\uff08Chebyshev Distance\uff09<\/h3>\n\n\n\n
import numpy as np\n\nx = np.array([1, 2, 3])\ny = np.array([4, 5, 6])\n\ndistance = np.max(np.abs(x - y))\nprint(f\"\u5207\u6bd4\u96ea\u592b\u8ddd\u79bb: {distance}\")\n<\/code><\/pre>\n\n\n\n5. \u4f59\u5f26\u76f8\u4f3c\u5ea6\uff08Cosine Similarity\uff09\u4e0e\u4f59\u5f26\u8ddd\u79bb<\/h3>\n\n\n\n
import numpy as np\nfrom sklearn.metrics.pairwise import cosine_similarity\n\nx = np.array([[1, 2, 3]])\ny = np.array([[4, 5, 6]])\n\nsimilarity = cosine_similarity(x, y)\ndistance = 1 - similarity[0][0]\nprint(f\"\u4f59\u5f26\u8ddd\u79bb: {distance}\")<\/code><\/pre>\n\n\n\n\u4e94\u3001\u5e94\u7528\u573a\u666f<\/h2>\n\n\n\n
1. \u5206\u7c7b\u95ee\u9898\u573a\u666f\uff1a<\/h3>\n\n\n\n
2. \u56de\u5f52\u95ee\u9898\u573a\u666f<\/h3>\n\n\n\n
\u516d\u3001\u5b9e\u6218\u6848\u4f8b<\/h2>\n\n\n\n
1. KNN \u5206\u7c7b\uff08iris\u6570\u636e\u96c6\uff09<\/h3>\n\n\n\n
\n
\n
from sklearn.datasets import load_iris\n\niris = load_iris()\n\nx = iris.data\n\ny = iris.target\n\nfeature_names = iris.feature_names\n\ntarget_names = iris.target_names<\/code><\/pre>\n\n\n\n![\"\"](\"https:\/\/zhangliangshan.cn\/wp-content\/uploads\/2025\/02\/image-27-1024x469.png\")
<\/div>from sklearn.datasets import load_iris\nfrom sklearn.model_selection import train_test_split\nfrom sklearn.neighbors import KNeighborsClassifier\nfrom sklearn.metrics import accuracy_score\n\n# \u52a0\u8f7d\u9e22\u5c3e\u82b1\u6570\u636e\u96c6\niris = load_iris()\nX = iris.data # \u7279\u5f81\u6570\u636e\ny = iris.target # \u6807\u7b7e\u6570\u636e\n\n# \u5212\u5206\u8bad\u7ec3\u96c6\u548c\u6d4b\u8bd5\u96c6\nX_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=42)\n\n# \u521b\u5efa KNN \u5206\u7c7b\u5668\uff0c\u6307\u5b9a\u8ddd\u79bb\u7b97\u6cd5\u4e3a\u66fc\u54c8\u987f\u8ddd\u79bb\nknn = KNeighborsClassifier(n_neighbors=5, metric='manhattan')\n\n# \u8bad\u7ec3\u6a21\u578b\nknn.fit(X_train, y_train)\n\n# \u9884\u6d4b\u548c\u8bc4\u4f30\ny_pred = knn.predict(X_test)\naccuracy = accuracy_score(y_test, y_pred)\nprint(f\"\u6a21\u578b\u7684\u51c6\u786e\u7387: {accuracy:.2f}\")\n\n# \u51c6\u5907\u51e0\u7ec4\u65b0\u6837\u672c\u7684\u7279\u5f81\u5411\u91cf\nnew_samples = [\n [5.1, 3.5, 1.4, 0.2],\n [6.2, 3.4, 5.4, 2.3],\n [5.7, 2.8, 4.1, 1.3]\n]\n\n# \u8fdb\u884c\u6279\u91cf\u9884\u6d4b\npredicted_classes = knn.predict(new_samples)\ntarget_names = iris.target_names\n\n# \u8f93\u51fa\u6bcf\u4e2a\u6837\u672c\u7684\u9884\u6d4b\u7c7b\u522b\u540d\u79f0\nfor i, predicted_class in enumerate(predicted_classes):\n predicted_class_name = target_names[predicted_class]\n print(f\"\u7b2c {i + 1} \u4e2a\u65b0\u6837\u672c\u7684\u9884\u6d4b\u7c7b\u522b\u662f: {predicted_class_name}\")<\/code><\/pre>\n\n\n\n![\"\"](\"https:\/\/zhangliangshan.cn\/wp-content\/uploads\/2025\/02\/image-28-1024x622.png\")
<\/div><\/figure>\n\n\n\n2. KNN\u56de\u5f52\u9884\u6d4b\uff08\u7cd6\u5c3f\u75c5\u6570\u636e\u96c6\uff09<\/h3>\n\n\n\n