\u56fe\u7247
\n\u6839\u636e\u6570\u636e\u7c7b\u578b\u7684\u4e0d\u540c\uff0c\u5bf9\u4e00\u4e2a\u95ee\u9898\u7684\u5efa\u6a21\u6709\u4e0d\u540c\u7684\u65b9\u5f0f\u3002\u5728\u673a\u5668\u5b66\u4e60\u6216\u8005\u4eba\u5de5\u667a\u80fd\u9886\u57df\uff0c\u4eba\u4eec\u9996\u5148\u4f1a\u8003\u8651\u7b97\u6cd5\u7684\u5b66\u4e60\u65b9\u5f0f\u3002\u5728\u673a\u5668\u5b66\u4e60\u9886\u57df\uff0c\u6709\u51e0\u79cd\u4e3b\u8981\u7684\u5b66\u4e60\u65b9\u5f0f\u3002\u5c06\u7b97\u6cd5\u6309\u7167\u5b66\u4e60\u65b9\u5f0f\u5206\u7c7b\u662f\u4e00\u4e2a\u4e0d\u9519\u7684\u60f3\u6cd5\uff0c\u8fd9\u6837\u53ef\u4ee5\u8ba9\u4eba\u4eec\u5728\u5efa\u6a21\u548c\u7b97\u6cd5\u9009\u62e9\u7684\u65f6\u5019\u8003\u8651\u80fd\u6839\u636e\u8f93\u5165\u6570\u636e\u6765\u9009\u62e9\u6700\u5408\u9002\u7684\u7b97\u6cd5\u6765\u83b7\u5f97\u6700\u597d\u7684\u7ed3\u679c\u3002<\/p>\n
<\/p>\n
1. \u76d1\u7763\u5f0f\u5b66\u4e60\uff1a<\/p>\n
\u56fe\u7247<\/p>\n
<\/p>\n
\u5728\u76d1\u7763\u5f0f\u5b66\u4e60\u4e0b\uff0c\u8f93\u5165\u6570\u636e\u88ab\u79f0\u4e3a\u201c\u8bad\u7ec3\u6570\u636e\u201d\uff0c\u6bcf\u7ec4\u8bad\u7ec3\u6570\u636e\u6709\u4e00\u4e2a\u660e\u786e\u7684\u6807\u8bc6\u6216\u7ed3\u679c\uff0c\u5982\u5bf9\u9632\u5783\u573e\u90ae\u4ef6\u7cfb\u7edf\u4e2d\u201c\u5783\u573e\u90ae\u4ef6\u201d\u201c\u975e\u5783\u573e\u90ae\u4ef6\u201d\uff0c\u5bf9\u624b\u5199\u6570\u5b57\u8bc6\u522b\u4e2d\u7684\u201c1\u201c\uff0c\u201d2\u201c\uff0c\u201d3\u201c\uff0c\u201d4\u201c\u7b49\u3002\u5728\u5efa\u7acb\u9884\u6d4b\u6a21\u578b\u7684\u65f6\u5019\uff0c\u76d1\u7763\u5f0f\u5b66\u4e60\u5efa\u7acb\u4e00\u4e2a\u5b66\u4e60\u8fc7\u7a0b\uff0c\u5c06\u9884\u6d4b\u7ed3\u679c\u4e0e\u201c\u8bad\u7ec3\u6570\u636e\u201d\u7684\u5b9e\u9645\u7ed3\u679c\u8fdb\u884c\u6bd4\u8f83\uff0c\u4e0d\u65ad\u7684\u8c03\u6574\u9884\u6d4b\u6a21\u578b\uff0c\u76f4\u5230\u6a21\u578b\u7684\u9884\u6d4b\u7ed3\u679c\u8fbe\u5230\u4e00\u4e2a\u9884\u671f\u7684\u51c6\u786e\u7387\u3002\u76d1\u7763\u5f0f\u5b66\u4e60\u7684\u5e38\u89c1\u5e94\u7528\u573a\u666f\u5982\u5206\u7c7b\u95ee\u9898\u548c\u56de\u5f52\u95ee\u9898\u3002\u5e38\u89c1\u7b97\u6cd5\u6709\u903b\u8f91\u56de\u5f52\uff08Logistic Regression\uff09\u548c\u53cd\u5411\u4f20\u9012\u795e\u7ecf\u7f51\u7edc\uff08Back Propagation Neural Network\uff09<\/p>\n
<\/p>\n
2. \u975e\u76d1\u7763\u5f0f\u5b66\u4e60\uff1a<\/p>\n
\u56fe\u7247<\/p>\n
<\/p>\n
\u5728\u975e\u76d1\u7763\u5f0f\u5b66\u4e60\u4e2d\uff0c\u6570\u636e\u5e76\u4e0d\u88ab\u7279\u522b\u6807\u8bc6\uff0c\u5b66\u4e60\u6a21\u578b\u662f\u4e3a\u4e86\u63a8\u65ad\u51fa\u6570\u636e\u7684\u4e00\u4e9b\u5185\u5728\u7ed3\u6784\u3002\u5e38\u89c1\u7684\u5e94\u7528\u573a\u666f\u5305\u62ec\u5173\u8054\u89c4\u5219\u7684\u5b66\u4e60\u4ee5\u53ca\u805a\u7c7b\u7b49\u3002\u5e38\u89c1\u7b97\u6cd5\u5305\u62ecApriori\u7b97\u6cd5\u4ee5\u53cak-Means\u7b97\u6cd5\u3002<\/p>\n
<\/p>\n
3. \u534a\u76d1\u7763\u5f0f\u5b66\u4e60\uff1a<\/p>\n
\u56fe\u7247<\/p>\n
<\/p>\n
\u5728\u6b64\u5b66\u4e60\u65b9\u5f0f\u4e0b\uff0c\u8f93\u5165\u6570\u636e\u90e8\u5206\u88ab\u6807\u8bc6\uff0c\u90e8\u5206\u6ca1\u6709\u88ab\u6807\u8bc6\uff0c\u8fd9\u79cd\u5b66\u4e60\u6a21\u578b\u53ef\u4ee5\u7528\u6765\u8fdb\u884c\u9884\u6d4b\uff0c\u4f46\u662f\u6a21\u578b\u9996\u5148\u9700\u8981\u5b66\u4e60\u6570\u636e\u7684\u5185\u5728\u7ed3\u6784\u4ee5\u4fbf\u5408\u7406\u7684\u7ec4\u7ec7\u6570\u636e\u6765\u8fdb\u884c\u9884\u6d4b\u3002\u5e94\u7528\u573a\u666f\u5305\u62ec\u5206\u7c7b\u548c\u56de\u5f52\uff0c\u7b97\u6cd5\u5305\u62ec\u4e00\u4e9b\u5bf9\u5e38\u7528\u76d1\u7763\u5f0f\u5b66\u4e60\u7b97\u6cd5\u7684\u5ef6\u4f38\uff0c\u8fd9\u4e9b\u7b97\u6cd5\u9996\u5148\u8bd5\u56fe\u5bf9\u672a\u6807\u8bc6\u6570\u636e\u8fdb\u884c\u5efa\u6a21\uff0c\u5728\u6b64\u57fa\u7840\u4e0a\u518d\u5bf9\u6807\u8bc6\u7684\u6570\u636e\u8fdb\u884c\u9884\u6d4b\u3002\u5982\u56fe\u8bba\u63a8\u7406\u7b97\u6cd5\uff08Graph Inference\uff09\u6216\u8005\u62c9\u666e\u62c9\u65af\u652f\u6301\u5411\u91cf\u673a\uff08Laplacian SVM.\uff09\u7b49\u3002<\/p>\n
<\/p>\n
4. \u5f3a\u5316\u5b66\u4e60\uff1a<\/p>\n
\u56fe\u7247<\/p>\n
<\/p>\n
\u5728\u8fd9\u79cd\u5b66\u4e60\u6a21\u5f0f\u4e0b\uff0c\u8f93\u5165\u6570\u636e\u4f5c\u4e3a\u5bf9\u6a21\u578b\u7684\u53cd\u9988\uff0c\u4e0d\u50cf\u76d1\u7763\u6a21\u578b\u90a3\u6837\uff0c\u8f93\u5165\u6570\u636e\u4ec5\u4ec5\u662f\u4f5c\u4e3a\u4e00\u4e2a\u68c0\u67e5\u6a21\u578b\u5bf9\u9519\u7684\u65b9\u5f0f\uff0c\u5728\u5f3a\u5316\u5b66\u4e60\u4e0b\uff0c\u8f93\u5165\u6570\u636e\u76f4\u63a5\u53cd\u9988\u5230\u6a21\u578b\uff0c\u6a21\u578b\u5fc5\u987b\u5bf9\u6b64\u7acb\u523b\u4f5c\u51fa\u8c03\u6574\u3002\u5e38\u89c1\u7684\u5e94\u7528\u573a\u666f\u5305\u62ec\u52a8\u6001\u7cfb\u7edf\u4ee5\u53ca\u673a\u5668\u4eba\u63a7\u5236\u7b49\u3002\u5e38\u89c1\u7b97\u6cd5\u5305\u62ecQ-Learning\u4ee5\u53ca\u65f6\u95f4\u5dee\u5b66\u4e60\uff08Temporal difference learning\uff09<\/p>\n
<\/p>\n
\u5728\u4f01\u4e1a\u6570\u636e\u5e94\u7528\u7684\u573a\u666f\u4e0b\uff0c \u4eba\u4eec\u6700\u5e38\u7528\u7684\u53ef\u80fd\u5c31\u662f\u76d1\u7763\u5f0f\u5b66\u4e60\u548c\u975e\u76d1\u7763\u5f0f\u5b66\u4e60\u7684\u6a21\u578b\u3002\u5728\u56fe\u50cf\u8bc6\u522b\u7b49\u9886\u57df\uff0c\u7531\u4e8e\u5b58\u5728\u5927\u91cf\u7684\u975e\u6807\u8bc6\u7684\u6570\u636e\u548c\u5c11\u91cf\u7684\u53ef\u6807\u8bc6\u6570\u636e\uff0c \u76ee\u524d\u534a\u76d1\u7763\u5f0f\u5b66\u4e60\u662f\u4e00\u4e2a\u5f88\u70ed\u7684\u8bdd\u9898\u3002\u800c\u5f3a\u5316\u5b66\u4e60\u66f4\u591a\u7684\u5e94\u7528\u5728\u673a\u5668\u4eba\u63a7\u5236\u53ca\u5176\u4ed6\u9700\u8981\u8fdb\u884c\u7cfb\u7edf\u63a7\u5236\u7684\u9886\u57df\u3002<\/p>\n
<\/p>\n
5. \u7b97\u6cd5\u7c7b\u4f3c\u6027<\/p>\n
\u6839\u636e\u7b97\u6cd5\u7684\u529f\u80fd\u548c\u5f62\u5f0f\u7684\u7c7b\u4f3c\u6027\uff0c\u6211\u4eec\u53ef\u4ee5\u628a\u7b97\u6cd5\u5206\u7c7b\uff0c\u6bd4\u5982\u8bf4\u57fa\u4e8e\u6811\u7684\u7b97\u6cd5\uff0c\u57fa\u4e8e\u795e\u7ecf\u7f51\u7edc\u7684\u7b97\u6cd5\u7b49\u7b49\u3002\u5f53\u7136\uff0c\u673a\u5668\u5b66\u4e60\u7684\u8303\u56f4\u975e\u5e38\u5e9e\u5927\uff0c\u6709\u4e9b\u7b97\u6cd5\u5f88\u96be\u660e\u786e\u5f52\u7c7b\u5230\u67d0\u4e00\u7c7b\u3002\u800c\u5bf9\u4e8e\u6709\u4e9b\u5206\u7c7b\u6765\u8bf4\uff0c\u540c\u4e00\u5206\u7c7b\u7684\u7b97\u6cd5\u53ef\u4ee5\u9488\u5bf9\u4e0d\u540c\u7c7b\u578b\u7684\u95ee\u9898\u3002\u8fd9\u91cc\uff0c\u6211\u4eec\u5c3d\u91cf\u628a\u5e38\u7528\u7684\u7b97\u6cd5\u6309\u7167\u6700\u5bb9\u6613\u7406\u89e3\u7684\u65b9\u5f0f\u8fdb\u884c\u5206\u7c7b\u3002<\/p>\n
<\/p>\n
6. \u56de\u5f52\u7b97\u6cd5\uff1a<\/p>\n
\u56fe\u7247<\/p>\n
<\/p>\n
\u56de\u5f52\u7b97\u6cd5\u662f\u8bd5\u56fe\u91c7\u7528\u5bf9\u8bef\u5dee\u7684\u8861\u91cf\u6765\u63a2\u7d22\u53d8\u91cf\u4e4b\u95f4\u7684\u5173\u7cfb\u7684\u4e00\u7c7b\u7b97\u6cd5\u3002\u56de\u5f52\u7b97\u6cd5\u662f\u7edf\u8ba1\u673a\u5668\u5b66\u4e60\u7684\u5229\u5668\u3002\u5728\u673a\u5668\u5b66\u4e60\u9886\u57df\uff0c\u4eba\u4eec\u8bf4\u8d77\u56de\u5f52\uff0c\u6709\u65f6\u5019\u662f\u6307\u4e00\u7c7b\u95ee\u9898\uff0c\u6709\u65f6\u5019\u662f\u6307\u4e00\u7c7b\u7b97\u6cd5\uff0c\u8fd9\u4e00\u70b9\u5e38\u5e38\u4f1a\u4f7f\u521d\u5b66\u8005\u6709\u6240\u56f0\u60d1\u3002\u5e38\u89c1\u7684\u56de\u5f52\u7b97\u6cd5\u5305\u62ec\uff1a\u6700\u5c0f\u4e8c\u4e58\u6cd5\uff08Ordinary Least Square\uff09\uff0c\u903b\u8f91\u56de\u5f52\uff08Logistic Regression\uff09\uff0c\u9010\u6b65\u5f0f\u56de\u5f52\uff08Stepwise Regression\uff09\uff0c\u591a\u5143\u81ea\u9002\u5e94\u56de\u5f52\u6837\u6761\uff08Multivariate Adaptive Regression Splines\uff09\u4ee5\u53ca\u672c\u5730\u6563\u70b9\u5e73\u6ed1\u4f30\u8ba1\uff08Locally Estimated Scatterplot Smoothing\uff09<\/p>\n
<\/p>\n
7. \u57fa\u4e8e\u5b9e\u4f8b\u7684\u7b97\u6cd5<\/p>\n
\u56fe\u7247<\/p>\n
<\/p>\n
\u57fa\u4e8e\u5b9e\u4f8b\u7684\u7b97\u6cd5\u5e38\u5e38\u7528\u6765\u5bf9\u51b3\u7b56\u95ee\u9898\u5efa\u7acb\u6a21\u578b\uff0c\u8fd9\u6837\u7684\u6a21\u578b\u5e38\u5e38\u5148\u9009\u53d6\u4e00\u6279\u6837\u672c\u6570\u636e\uff0c\u7136\u540e\u6839\u636e\u67d0\u4e9b\u8fd1\u4f3c\u6027\u628a\u65b0\u6570\u636e\u4e0e\u6837\u672c\u6570\u636e\u8fdb\u884c\u6bd4\u8f83\u3002\u901a\u8fc7\u8fd9\u79cd\u65b9\u5f0f\u6765\u5bfb\u627e\u6700\u4f73\u7684\u5339\u914d\u3002\u56e0\u6b64\uff0c\u57fa\u4e8e\u5b9e\u4f8b\u7684\u7b97\u6cd5\u5e38\u5e38\u4e5f\u88ab\u79f0\u4e3a\u201c\u8d62\u5bb6\u901a\u5403\u201d\u5b66\u4e60\u6216\u8005\u201c\u57fa\u4e8e\u8bb0\u5fc6\u7684\u5b66\u4e60\u201d\u3002\u5e38\u89c1\u7684\u7b97\u6cd5\u5305\u62ec k-Nearest Neighbor(KNN), \u5b66\u4e60\u77e2\u91cf\u91cf\u5316\uff08Learning Vector Quantization\uff0c LVQ\uff09\uff0c\u4ee5\u53ca\u81ea\u7ec4\u7ec7\u6620\u5c04\u7b97\u6cd5\uff08Self-Organizing Map \uff0c SOM\uff09<\/p>\n
<\/p>\n
8. \u6b63\u5219\u5316\u65b9\u6cd5<\/p>\n
\u56fe\u7247<\/p>\n
<\/p>\n
\u6b63\u5219\u5316\u65b9\u6cd5\u662f\u5176\u4ed6\u7b97\u6cd5\uff08\u901a\u5e38\u662f\u56de\u5f52\u7b97\u6cd5\uff09\u7684\u5ef6\u4f38\uff0c\u6839\u636e\u7b97\u6cd5\u7684\u590d\u6742\u5ea6\u5bf9\u7b97\u6cd5\u8fdb\u884c\u8c03\u6574\u3002\u6b63\u5219\u5316\u65b9\u6cd5\u901a\u5e38\u5bf9\u7b80\u5355\u6a21\u578b\u4e88\u4ee5\u5956\u52b1\u800c\u5bf9\u590d\u6742\u7b97\u6cd5\u4e88\u4ee5\u60e9\u7f5a\u3002\u5e38\u89c1\u7684\u7b97\u6cd5\u5305\u62ec\uff1aRidge Regression\uff0cLeast Absolute Shrinkage and Selection Operator\uff08LASSO\uff09\uff0c\u4ee5\u53ca\u5f39\u6027\u7f51\u7edc\uff08Elastic Net\uff09\u3002<\/p>\n
<\/p>\n
9. \u51b3\u7b56\u6811\u5b66\u4e60<\/p>\n
\u56fe\u7247<\/p>\n
<\/p>\n
\u51b3\u7b56\u6811\u7b97\u6cd5\u6839\u636e\u6570\u636e\u7684\u5c5e\u6027\u91c7\u7528\u6811\u72b6\u7ed3\u6784\u5efa\u7acb\u51b3\u7b56\u6a21\u578b\uff0c \u51b3\u7b56\u6811\u6a21\u578b\u5e38\u5e38\u7528\u6765\u89e3\u51b3\u5206\u7c7b\u548c\u56de\u5f52\u95ee\u9898\u3002\u5e38\u89c1\u7684\u7b97\u6cd5\u5305\u62ec\uff1a\u5206\u7c7b\u53ca\u56de\u5f52\u6811\uff08Classification And Regression Tree\uff0c CART\uff09\uff0c ID3 (Iterative Dichotomiser 3)\uff0c C4.5\uff0c Chi-squared Automatic Interaction Detection(CHAID), Decision Stump, \u968f\u673a\u68ee\u6797\uff08Random Forest\uff09\uff0c \u591a\u5143\u81ea\u9002\u5e94\u56de\u5f52\u6837\u6761\uff08MARS\uff09\u4ee5\u53ca\u68af\u5ea6\u63a8\u8fdb\u673a\uff08Gradient Boosting Machine\uff0c GBM\uff09<\/p>\n
<\/p>\n
10. \u8d1d\u53f6\u65af\u65b9\u6cd5<\/p>\n
\u56fe\u7247<\/p>\n
<\/p>\n
\u8d1d\u53f6\u65af\u65b9\u6cd5\u7b97\u6cd5\u662f\u57fa\u4e8e\u8d1d\u53f6\u65af\u5b9a\u7406\u7684\u4e00\u7c7b\u7b97\u6cd5\uff0c\u4e3b\u8981\u7528\u6765\u89e3\u51b3\u5206\u7c7b\u548c\u56de\u5f52\u95ee\u9898\u3002\u5e38\u89c1\u7b97\u6cd5\u5305\u62ec\uff1a\u6734\u7d20\u8d1d\u53f6\u65af\u7b97\u6cd5\uff0c\u5e73\u5747\u5355\u4f9d\u8d56\u4f30\u8ba1\uff08Averaged One-Dependence Estimators\uff0c AODE\uff09\uff0c\u4ee5\u53caBayesian Belief Network\uff08BBN\uff09\u3002<\/p>\n
<\/p>\n
11. \u57fa\u4e8e\u6838\u7684\u7b97\u6cd5<\/p>\n
\u56fe\u7247<\/p>\n
<\/p>\n
\u57fa\u4e8e\u6838\u7684\u7b97\u6cd5\u4e2d\u6700\u8457\u540d\u7684\u83ab\u8fc7\u4e8e\u652f\u6301\u5411\u91cf\u673a\uff08SVM\uff09\u4e86\u3002\u57fa\u4e8e\u6838\u7684\u7b97\u6cd5\u628a\u8f93\u5165\u6570\u636e\u6620\u5c04\u5230\u4e00\u4e2a\u9ad8\u9636\u7684\u5411\u91cf\u7a7a\u95f4\uff0c \u5728\u8fd9\u4e9b\u9ad8\u9636\u5411\u91cf\u7a7a\u95f4\u91cc\uff0c \u6709\u4e9b\u5206\u7c7b\u6216\u8005\u56de\u5f52\u95ee\u9898\u80fd\u591f\u66f4\u5bb9\u6613\u7684\u89e3\u51b3\u3002\u5e38\u89c1\u7684\u57fa\u4e8e\u6838\u7684\u7b97\u6cd5\u5305\u62ec\uff1a\u652f\u6301\u5411\u91cf\u673a\uff08Support Vector Machine\uff0c SVM\uff09\uff0c \u5f84\u5411\u57fa\u51fd\u6570\uff08Radial Basis Function \uff0cRBF)\uff0c \u4ee5\u53ca\u7ebf\u6027\u5224\u522b\u5206\u6790\uff08Linear Discriminate Analysis \uff0cLDA)\u7b49<\/p>\n
<\/p>\n
12.\u805a\u7c7b\u7b97\u6cd5<\/p>\n
\u56fe\u7247<\/p>\n
\u805a\u7c7b\uff0c\u5c31\u50cf\u56de\u5f52\u4e00\u6837\uff0c\u6709\u65f6\u5019\u4eba\u4eec\u63cf\u8ff0\u7684\u662f\u4e00\u7c7b\u95ee\u9898\uff0c\u6709\u65f6\u5019\u63cf\u8ff0\u7684\u662f\u4e00\u7c7b\u7b97\u6cd5\u3002\u805a\u7c7b\u7b97\u6cd5\u901a\u5e38\u6309\u7167\u4e2d\u5fc3\u70b9\u6216\u8005\u5206\u5c42\u7684\u65b9\u5f0f\u5bf9\u8f93\u5165\u6570\u636e\u8fdb\u884c\u5f52\u5e76\u3002\u6240\u4ee5\u7684\u805a\u7c7b\u7b97\u6cd5\u90fd\u8bd5\u56fe\u627e\u5230\u6570\u636e\u7684\u5185\u5728\u7ed3\u6784\uff0c\u4ee5\u4fbf\u6309\u7167\u6700\u5927\u7684\u5171\u540c\u70b9\u5c06\u6570\u636e\u8fdb\u884c\u5f52\u7c7b\u3002\u5e38\u89c1\u7684\u805a\u7c7b\u7b97\u6cd5\u5305\u62ec k-Means\u7b97\u6cd5\u4ee5\u53ca\u671f\u671b\u6700\u5927\u5316\u7b97\u6cd5\uff08Expectation Maximization\uff0c EM\uff09\u3002<\/p>\n
<\/p>\n
13. \u5173\u8054\u89c4\u5219\u5b66\u4e60<\/p>\n
\u56fe\u7247<\/p>\n
\u5173\u8054\u89c4\u5219\u5b66\u4e60\u901a\u8fc7\u5bfb\u627e\u6700\u80fd\u591f\u89e3\u91ca\u6570\u636e\u53d8\u91cf\u4e4b\u95f4\u5173\u7cfb\u7684\u89c4\u5219\uff0c\u6765\u627e\u51fa\u5927\u91cf\u591a\u5143\u6570\u636e\u96c6\u4e2d\u6709\u7528\u7684\u5173\u8054\u89c4\u5219\u3002\u5e38\u89c1\u7b97\u6cd5\u5305\u62ec Apriori\u7b97\u6cd5\u548cEclat\u7b97\u6cd5\u7b49\u3002<\/p>\n
<\/p>\n
14. \u4eba\u5de5\u795e\u7ecf\u7f51\u7edc<\/p>\n
\u56fe\u7247<\/p>\n
<\/p>\n
\u4eba\u5de5\u795e\u7ecf\u7f51\u7edc\u7b97\u6cd5\u6a21\u62df\u751f\u7269\u795e\u7ecf\u7f51\u7edc\uff0c\u662f\u4e00\u7c7b\u6a21\u5f0f\u5339\u914d\u7b97\u6cd5\u3002\u901a\u5e38\u7528\u4e8e\u89e3\u51b3\u5206\u7c7b\u548c\u56de\u5f52\u95ee\u9898\u3002\u4eba\u5de5\u795e\u7ecf\u7f51\u7edc\u662f\u673a\u5668\u5b66\u4e60\u7684\u4e00\u4e2a\u5e9e\u5927\u7684\u5206\u652f\uff0c\u6709\u51e0\u767e\u79cd\u4e0d\u540c\u7684\u7b97\u6cd5\u3002\uff08\u5176\u4e2d\u6df1\u5ea6\u5b66\u4e60\u5c31\u662f\u5176\u4e2d\u7684\u4e00\u7c7b\u7b97\u6cd5\uff0c\u6211\u4eec\u4f1a\u5355\u72ec\u8ba8\u8bba\uff09\uff0c\u91cd\u8981\u7684\u4eba\u5de5\u795e\u7ecf\u7f51\u7edc\u7b97\u6cd5\u5305\u62ec\uff1a\u611f\u77e5\u5668\u795e\u7ecf\u7f51\u7edc\uff08Perceptron Neural Network\uff09, \u53cd\u5411\u4f20\u9012\uff08Back Propagation\uff09\uff0c Hopfield\u7f51\u7edc\uff0c\u81ea\u7ec4\u7ec7\u6620\u5c04\uff08Self-Organizing Map, SOM\uff09\u3002\u5b66\u4e60\u77e2\u91cf\u91cf\u5316\uff08Learning Vector Quantization\uff0c LVQ\uff09<\/p>\n
<\/p>\n
15. \u6df1\u5ea6\u5b66\u4e60<\/p>\n
\u56fe\u7247<\/p>\n
<\/p>\n
\u6df1\u5ea6\u5b66\u4e60\u7b97\u6cd5\u662f\u5bf9\u4eba\u5de5\u795e\u7ecf\u7f51\u7edc\u7684\u53d1\u5c55\u3002\u5728\u8fd1\u671f\u8d62\u5f97\u4e86\u5f88\u591a\u5173\u6ce8\uff0c \u7279\u522b\u662f\u767e\u5ea6\u4e5f\u5f00\u59cb\u53d1\u529b\u6df1\u5ea6\u5b66\u4e60\u540e\uff0c \u66f4\u662f\u5728\u56fd\u5185\u5f15\u8d77\u4e86\u5f88\u591a\u5173\u6ce8\u3002 \u5728\u8ba1\u7b97\u80fd\u529b\u53d8\u5f97\u65e5\u76ca\u5ec9\u4ef7\u7684\u4eca\u5929\uff0c\u6df1\u5ea6\u5b66\u4e60\u8bd5\u56fe\u5efa\u7acb\u5927\u5f97\u591a\u4e5f\u590d\u6742\u5f97\u591a\u7684\u795e\u7ecf\u7f51\u7edc\u3002\u5f88\u591a\u6df1\u5ea6\u5b66\u4e60\u7684\u7b97\u6cd5\u662f\u534a\u76d1\u7763\u5f0f\u5b66\u4e60\u7b97\u6cd5\uff0c\u7528\u6765\u5904\u7406\u5b58\u5728\u5c11\u91cf\u672a\u6807\u8bc6\u6570\u636e\u7684\u5927\u6570\u636e\u96c6\u3002\u5e38\u89c1\u7684\u6df1\u5ea6\u5b66\u4e60\u7b97\u6cd5\u5305\u62ec\uff1a\u53d7\u9650\u6ce2\u5c14\u5179\u66fc\u673a\uff08Restricted Boltzmann Machine\uff0c RBN\uff09\uff0c Deep Belief Networks\uff08DBN\uff09\uff0c\u5377\u79ef\u7f51\u7edc\uff08Convolutional Network\uff09, \u5806\u6808\u5f0f\u81ea\u52a8\u7f16\u7801\u5668\uff08Stacked Auto-encoders\uff09\u3002<\/p>\n
<\/p>\n
16. \u964d\u4f4e\u7ef4\u5ea6\u7b97\u6cd5<\/p>\n
\u56fe\u7247<\/p>\n
<\/p>\n
\u50cf\u805a\u7c7b\u7b97\u6cd5\u4e00\u6837\uff0c\u964d\u4f4e\u7ef4\u5ea6\u7b97\u6cd5\u8bd5\u56fe\u5206\u6790\u6570\u636e\u7684\u5185\u5728\u7ed3\u6784\uff0c\u4e0d\u8fc7\u964d\u4f4e\u7ef4\u5ea6\u7b97\u6cd5\u662f\u4ee5\u975e\u76d1\u7763\u5b66\u4e60\u7684\u65b9\u5f0f\u8bd5\u56fe\u5229\u7528\u8f83\u5c11\u7684\u4fe1\u606f\u6765\u5f52\u7eb3\u6216\u8005\u89e3\u91ca\u6570\u636e\u3002\u8fd9\u7c7b\u7b97\u6cd5\u53ef\u4ee5\u7528\u4e8e\u9ad8\u7ef4\u6570\u636e\u7684\u53ef\u89c6\u5316\u6216\u8005\u7528\u6765\u7b80\u5316\u6570\u636e\u4ee5\u4fbf\u76d1\u7763\u5f0f\u5b66\u4e60\u4f7f\u7528\u3002<\/p>\n
<\/p>\n
\u5e38\u89c1\u7684\u7b97\u6cd5\u5305\u62ec\uff1a\u4e3b\u6210\u4efd\u5206\u6790\uff08Principle Component Analysis\uff0c PCA\uff09\uff0c\u504f\u6700\u5c0f\u4e8c\u4e58\u56de\u5f52\uff08Partial Least Square Regression\uff0cPLS\uff09\uff0c Sammon\u6620\u5c04\uff0c\u591a\u7ef4\u5c3a\u5ea6\uff08Multi-Dimensional Scaling, MDS\uff09, \u6295\u5f71\u8ffd\u8e2a\uff08Projection Pursuit\uff09\u7b49\u3002<\/p>\n
<\/p>\n
17. \u96c6\u6210\u7b97\u6cd5\uff1a<\/p>\n
\u56fe\u7247<\/p>\n
<\/p>\n
\u96c6\u6210\u7b97\u6cd5\u7528\u4e00\u4e9b\u76f8\u5bf9\u8f83\u5f31\u7684\u5b66\u4e60\u6a21\u578b\u72ec\u7acb\u5730\u5c31\u540c\u6837\u7684\u6837\u672c\u8fdb\u884c\u8bad\u7ec3\uff0c\u7136\u540e\u628a\u7ed3\u679c\u6574\u5408\u8d77\u6765\u8fdb\u884c\u6574\u4f53\u9884\u6d4b\u3002\u96c6\u6210\u7b97\u6cd5\u7684\u4e3b\u8981\u96be\u70b9\u5728\u4e8e\u7a76\u7adf\u96c6\u6210\u54ea\u4e9b\u72ec\u7acb\u7684\u8f83\u5f31\u7684\u5b66\u4e60\u6a21\u578b\u4ee5\u53ca\u5982\u4f55\u628a\u5b66\u4e60\u7ed3\u679c\u6574\u5408\u8d77\u6765\u3002<\/p>\n
<\/p>\n
\u8fd9\u662f\u4e00\u7c7b\u975e\u5e38\u5f3a\u5927\u7684\u7b97\u6cd5\uff0c\u540c\u65f6\u4e5f\u975e\u5e38\u6d41\u884c\u3002\u5e38\u89c1\u7684\u7b97\u6cd5\u5305\u62ec\uff1aBoosting\uff0c Bootstrapped Aggregation\uff08Bagging\uff09\uff0c AdaBoost\uff0c\u5806\u53e0\u6cdb\u5316\uff08Stacked Generalization\uff0c Blending\uff09\uff0c\u68af\u5ea6\u63a8\u8fdb\u673a\uff08Gradient Boosting Machine, GBM\uff09\uff0c\u968f\u673a\u68ee\u6797\uff08Random Forest\uff09\u3002<\/p>\n
<\/p>\n
\u5e38\u89c1\u673a\u5668\u5b66\u4e60\u7b97\u6cd5\u4f18\u7f3a\u70b9\uff1a<\/p>\n
<\/p>\n
\u6734\u7d20\u8d1d\u53f6\u65af\uff1a<\/p>\n
1. \u5982\u679c\u7ed9\u51fa\u7684\u7279\u5f81\u5411\u91cf\u957f\u5ea6\u53ef\u80fd\u4e0d\u540c\uff0c\u8fd9\u662f\u9700\u8981\u5f52\u4e00\u5316\u4e3a\u901a\u957f\u5ea6\u7684\u5411\u91cf\uff08\u8fd9\u91cc\u4ee5\u6587\u672c\u5206\u7c7b\u4e3a\u4f8b\uff09\uff0c\u6bd4\u5982\u8bf4\u662f\u53e5\u5b50\u5355\u8bcd\u7684\u8bdd\uff0c\u5219\u957f\u5ea6\u4e3a\u6574\u4e2a\u8bcd\u6c47\u91cf\u7684\u957f\u5ea6\uff0c\u5bf9\u5e94\u4f4d\u7f6e\u662f\u8be5\u5355\u8bcd\u51fa\u73b0\u7684\u6b21\u6570\u3002<\/p>\n
2. \u8ba1\u7b97\u516c\u5f0f\u5982\u4e0b\uff1a<\/p>\n
\u56fe\u7247<\/p>\n
\u5176\u4e2d\u4e00\u9879\u6761\u4ef6\u6982\u7387\u53ef\u4ee5\u901a\u8fc7\u6734\u7d20\u8d1d\u53f6\u65af\u6761\u4ef6\u72ec\u7acb\u5c55\u5f00\u3002\u8981\u6ce8\u610f\u4e00\u70b9\u5c31\u662f \u56fe\u7247\u7684\u8ba1\u7b97\u65b9\u6cd5\uff0c\u800c\u7531\u6734\u7d20\u8d1d\u53f6\u65af\u7684\u524d\u63d0\u5047\u8bbe\u53ef\u77e5\uff0c\u56fe\u7247 =\u56fe\u7247 \uff0c\u56e0\u6b64\u4e00\u822c\u6709\u4e24\u79cd\uff0c\u4e00\u79cd\u662f\u5728\u7c7b\u522b\u4e3aci\u7684\u90a3\u4e9b\u6837\u672c\u96c6\u4e2d\uff0c\u627e\u5230wj\u51fa\u73b0\u6b21\u6570\u7684\u603b\u548c\uff0c\u7136\u540e\u9664\u4ee5\u8be5\u6837\u672c\u7684\u603b\u548c\uff1b\u7b2c\u4e8c\u79cd\u65b9\u6cd5\u662f\u7c7b\u522b\u4e3aci\u7684\u90a3\u4e9b\u6837\u672c\u96c6\u4e2d\uff0c\u627e\u5230wj\u51fa\u73b0\u6b21\u6570\u7684\u603b\u548c\uff0c\u7136\u540e\u9664\u4ee5\u8be5\u6837\u672c\u4e2d\u6240\u6709\u7279\u5f81\u51fa\u73b0\u6b21\u6570\u7684\u603b\u548c\u3002<\/p>\n
<\/p>\n
3. \u5982\u679c \u56fe\u7247\u4e2d\u7684\u67d0\u4e00\u9879\u4e3a0\uff0c\u5219\u5176\u8054\u5408\u6982\u7387\u7684\u4e58\u79ef\u4e5f\u53ef\u80fd\u4e3a0\uff0c\u53732\u4e2d\u516c\u5f0f\u7684\u5206\u5b50\u4e3a0\uff0c\u4e3a\u4e86\u907f\u514d\u8fd9\u79cd\u73b0\u8c61\u51fa\u73b0\uff0c\u4e00\u822c\u60c5\u51b5\u4e0b\u4f1a\u5c06\u8fd9\u4e00\u9879\u521d\u59cb\u5316\u4e3a1\uff0c\u5f53\u7136\u4e3a\u4e86\u4fdd\u8bc1\u6982\u7387\u76f8\u7b49\uff0c\u5206\u6bcd\u5e94\u5bf9\u5e94\u521d\u59cb\u5316\u4e3a2\uff08\u8fd9\u91cc\u56e0\u4e3a\u662f2\u7c7b\uff0c\u6240\u4ee5\u52a02\uff0c\u5982\u679c\u662fk\u7c7b\u5c31\u9700\u8981\u52a0k\uff0c\u672f\u8bed\u4e0a\u53eb\u505alaplace\u5149\u6ed1, \u5206\u6bcd\u52a0k\u7684\u539f\u56e0\u662f\u4f7f\u4e4b\u6ee1\u8db3\u5168\u6982\u7387\u516c\u5f0f\uff09\u3002<\/p>\n
<\/p>\n
\u6734\u7d20\u8d1d\u53f6\u65af\u7684\u4f18\u70b9\uff1a\u5bf9\u5c0f\u89c4\u6a21\u7684\u6570\u636e\u8868\u73b0\u5f88\u597d\uff0c\u9002\u5408\u591a\u5206\u7c7b\u4efb\u52a1\uff0c\u9002\u5408\u589e\u91cf\u5f0f\u8bad\u7ec3\u3002<\/p>\n
\u7f3a\u70b9\uff1a\u5bf9\u8f93\u5165\u6570\u636e\u7684\u8868\u8fbe\u5f62\u5f0f\u5f88\u654f\u611f\u3002<\/p>\n
<\/p>\n
\u51b3\u7b56\u6811\uff1a\u51b3\u7b56\u6811\u4e2d\u5f88\u91cd\u8981\u7684\u4e00\u70b9\u5c31\u662f\u9009\u62e9\u4e00\u4e2a\u5c5e\u6027\u8fdb\u884c\u5206\u679d\uff0c\u56e0\u6b64\u8981\u6ce8\u610f\u4e00\u4e0b\u4fe1\u606f\u589e\u76ca\u7684\u8ba1\u7b97\u516c\u5f0f\uff0c\u5e76\u6df1\u5165\u7406\u89e3\u5b83\u3002<\/p>\n
<\/p>\n
\u4fe1\u606f\u71b5\u7684\u8ba1\u7b97\u516c\u5f0f\u5982\u4e0b:<\/p>\n
\u56fe\u7247<\/p>\n
\u5176\u4e2d\u7684n\u4ee3\u8868\u6709n\u4e2a\u5206\u7c7b\u7c7b\u522b\uff08\u6bd4\u5982\u5047\u8bbe\u662f2\u7c7b\u95ee\u9898\uff0c\u90a3\u4e48n=2\uff09\u3002\u5206\u522b\u8ba1\u7b97\u8fd92\u7c7b\u6837\u672c\u5728\u603b\u6837\u672c\u4e2d\u51fa\u73b0\u7684\u6982\u7387p1\u548cp2\uff0c\u8fd9\u6837\u5c31\u53ef\u4ee5\u8ba1\u7b97\u51fa\u672a\u9009\u4e2d\u5c5e\u6027\u5206\u679d\u524d\u7684\u4fe1\u606f\u71b5\u3002<\/p>\n
<\/p>\n
\u73b0\u5728\u9009\u4e2d\u4e00\u4e2a\u5c5e\u6027xi\u7528\u6765\u8fdb\u884c\u5206\u679d\uff0c\u6b64\u65f6\u5206\u679d\u89c4\u5219\u662f\uff1a\u5982\u679cxi=vx\u7684\u8bdd\uff0c\u5c06\u6837\u672c\u5206\u5230\u6811\u7684\u4e00\u4e2a\u5206\u652f\uff1b\u5982\u679c\u4e0d\u76f8\u7b49\u5219\u8fdb\u5165\u53e6\u4e00\u4e2a\u5206\u652f\u3002\u5f88\u663e\u7136\uff0c\u5206\u652f\u4e2d\u7684\u6837\u672c\u5f88\u6709\u53ef\u80fd\u5305\u62ec2\u4e2a\u7c7b\u522b\uff0c\u5206\u522b\u8ba1\u7b97\u8fd92\u4e2a\u5206\u652f\u7684\u71b5H1\u548cH2,\u8ba1\u7b97\u51fa\u5206\u679d\u540e\u7684\u603b\u4fe1\u606f\u71b5H\u2019=p1*H1+p2*H2.\uff0c\u5219\u6b64\u65f6\u7684\u4fe1\u606f\u589e\u76ca\u0394H=H-H\u2019\u3002\u4ee5\u4fe1\u606f\u589e\u76ca\u4e3a\u539f\u5219\uff0c\u628a\u6240\u6709\u7684\u5c5e\u6027\u90fd\u6d4b\u8bd5\u4e00\u8fb9\uff0c\u9009\u62e9\u4e00\u4e2a\u4f7f\u589e\u76ca\u6700\u5927\u7684\u5c5e\u6027\u4f5c\u4e3a\u672c\u6b21\u5206\u679d\u5c5e\u6027\u3002<\/p>\n
<\/p>\n
\u51b3\u7b56\u6811\u7684\u4f18\u70b9\uff1a\u8ba1\u7b97\u91cf\u7b80\u5355\uff0c\u53ef\u89e3\u91ca\u6027\u5f3a\uff0c\u6bd4\u8f83\u9002\u5408\u5904\u7406\u6709\u7f3a\u5931\u5c5e\u6027\u503c\u7684\u6837\u672c\uff0c\u80fd\u591f\u5904\u7406\u4e0d\u76f8\u5173\u7684\u7279\u5f81\uff1b<\/p>\n
\u7f3a\u70b9\uff1a\u5bb9\u6613\u8fc7\u62df\u5408\uff08\u540e\u7eed\u51fa\u73b0\u4e86\u968f\u673a\u68ee\u6797\uff0c\u51cf\u5c0f\u4e86\u8fc7\u62df\u5408\u73b0\u8c61\uff09\u3002<\/p>\n
<\/p>\n
Logistic\u56de\u5f52\uff1aLogistic\u662f\u7528\u6765\u5206\u7c7b\u7684\uff0c\u662f\u4e00\u79cd\u7ebf\u6027\u5206\u7c7b\u5668\uff0c\u9700\u8981\u6ce8\u610f\u7684\u5730\u65b9\u6709\uff1a<\/p>\n
<\/p>\n
1. logistic\u51fd\u6570\u8868\u8fbe\u5f0f\u4e3a\uff1a<\/p>\n
\u56fe\u7247<\/p>\n
\u5176\u5bfc\u6570\u5f62\u5f0f\u4e3a\uff1a<\/p>\n
\u56fe\u7247<\/p>\n
<\/p>\n
2. logsitc\u56de\u5f52\u65b9\u6cd5\u4e3b\u8981\u662f\u7528\u6700\u5927\u4f3c\u7136\u4f30\u8ba1\u6765\u5b66\u4e60\u7684\uff0c\u6240\u4ee5\u5355\u4e2a\u6837\u672c\u7684\u540e\u9a8c\u6982\u7387\u4e3a\uff1a<\/p>\n
\u56fe\u7247<\/p>\n
\u5230\u6574\u4e2a\u6837\u672c\u7684\u540e\u9a8c\u6982\u7387\uff1a<\/p>\n
\u56fe\u7247<\/p>\n
\u5176\u4e2d\uff1a<\/p>\n
\u56fe\u7247<\/p>\n
\u901a\u8fc7\u5bf9\u6570\u8fdb\u4e00\u6b65\u5316\u7b80\u4e3a\uff1a<\/p>\n
<\/p>\n
\u56fe\u7247<\/p>\n
<\/p>\n
3. \u5176\u5b9e\u5b83\u7684loss function\u4e3a-l(\u03b8)\uff0c\u56e0\u6b64\u6211\u4eec\u9700\u4f7floss function\u6700\u5c0f\uff0c\u53ef\u91c7\u7528\u68af\u5ea6\u4e0b\u964d\u6cd5\u5f97\u5230\u3002\u68af\u5ea6\u4e0b\u964d\u6cd5\u516c\u5f0f\u4e3a:<\/p>\n
<\/p>\n
\u56fe\u7247<\/p>\n
\u56fe\u7247<\/p>\n
Logistic\u56de\u5f52\u4f18\u70b9\uff1a<\/p>\n
1. \u5b9e\u73b0\u7b80\u5355<\/p>\n
2. \u5206\u7c7b\u65f6\u8ba1\u7b97\u91cf\u975e\u5e38\u5c0f\uff0c\u901f\u5ea6\u5f88\u5feb\uff0c\u5b58\u50a8\u8d44\u6e90\u4f4e\uff1b<\/p>\n
<\/p>\n
\u7f3a\u70b9\uff1a<\/p>\n
1. \u5bb9\u6613\u6b20\u62df\u5408\uff0c\u4e00\u822c\u51c6\u786e\u5ea6\u4e0d\u592a\u9ad8<\/p>\n
2. \u53ea\u80fd\u5904\u7406\u4e24\u5206\u7c7b\u95ee\u9898\uff08\u5728\u6b64\u57fa\u7840\u4e0a\u884d\u751f\u51fa\u6765\u7684softmax\u53ef\u4ee5\u7528\u4e8e\u591a\u5206\u7c7b\uff09\uff0c\u4e14\u5fc5\u987b\u7ebf\u6027\u53ef\u5206\uff1b<\/p>\n
<\/p>\n
\u7ebf\u6027\u56de\u5f52\uff1a<\/p>\n
\u7ebf\u6027\u56de\u5f52\u624d\u662f\u771f\u6b63\u7528\u4e8e\u56de\u5f52\u7684\uff0c\u800c\u4e0d\u50cflogistic\u56de\u5f52\u662f\u7528\u4e8e\u5206\u7c7b\uff0c\u5176\u57fa\u672c\u601d\u60f3\u662f\u7528\u68af\u5ea6\u4e0b\u964d\u6cd5\u5bf9\u6700\u5c0f\u4e8c\u4e58\u6cd5\u5f62\u5f0f\u7684\u8bef\u5dee\u51fd\u6570\u8fdb\u884c\u4f18\u5316\uff0c\u5f53\u7136\u4e5f\u53ef\u4ee5\u7528normal equation\u76f4\u63a5\u6c42\u5f97\u53c2\u6570\u7684\u89e3\uff0c\u7ed3\u679c\u4e3a\uff1a<\/p>\n
<\/p>\n
\u56fe\u7247<\/p>\n
\u800c\u5728LWLR\uff08\u5c40\u90e8\u52a0\u6743\u7ebf\u6027\u56de\u5f52\uff09\u4e2d\uff0c\u53c2\u6570\u7684\u8ba1\u7b97\u8868\u8fbe\u5f0f\u4e3a:<\/p>\n
\u56fe\u7247<\/p>\n
\u56e0\u4e3a\u6b64\u65f6\u4f18\u5316\u7684\u662f\uff1a<\/p>\n
\u56fe\u7247<\/p>\n
\u7531\u6b64\u53ef\u89c1LWLR\u4e0eLR\u4e0d\u540c\uff0cLWLR\u662f\u4e00\u4e2a\u975e\u53c2\u6570\u6a21\u578b\uff0c\u56e0\u4e3a\u6bcf\u6b21\u8fdb\u884c\u56de\u5f52\u8ba1\u7b97\u90fd\u8981\u904d\u5386\u8bad\u7ec3\u6837\u672c\u81f3\u5c11\u4e00\u6b21\u3002<\/p>\n
<\/p>\n
\u7ebf\u6027\u56de\u5f52\u4f18\u70b9\uff1a\u5b9e\u73b0\u7b80\u5355\uff0c\u8ba1\u7b97\u7b80\u5355\uff1b<\/p>\n
\u7f3a\u70b9\uff1a\u4e0d\u80fd\u62df\u5408\u975e\u7ebf\u6027\u6570\u636e\uff1b<\/p>\n
<\/p>\n
KNN\u7b97\u6cd5\uff1aKNN\u5373\u6700\u8fd1\u90bb\u7b97\u6cd5\uff0c\u5176\u4e3b\u8981\u8fc7\u7a0b\u4e3a\uff1a<\/p>\n
1. \u8ba1\u7b97\u8bad\u7ec3\u6837\u672c\u548c\u6d4b\u8bd5\u6837\u672c\u4e2d\u6bcf\u4e2a\u6837\u672c\u70b9\u7684\u8ddd\u79bb\uff08\u5e38\u89c1\u7684\u8ddd\u79bb\u5ea6\u91cf\u6709\u6b27\u5f0f\u8ddd\u79bb\uff0c\u9a6c\u6c0f\u8ddd\u79bb\u7b49\uff09\uff1b<\/p>\n
2. \u5bf9\u4e0a\u9762\u6240\u6709\u7684\u8ddd\u79bb\u503c\u8fdb\u884c\u6392\u5e8f\uff1b<\/p>\n
3. \u9009\u524dk\u4e2a\u6700\u5c0f\u8ddd\u79bb\u7684\u6837\u672c\uff1b<\/p>\n
4. \u6839\u636e\u8fd9k\u4e2a\u6837\u672c\u7684\u6807\u7b7e\u8fdb\u884c\u6295\u7968\uff0c\u5f97\u5230\u6700\u540e\u7684\u5206\u7c7b\u7c7b\u522b\uff1b<\/p>\n
<\/p>\n
\u5982\u4f55\u9009\u62e9\u4e00\u4e2a\u6700\u4f73\u7684K\u503c\uff0c\u8fd9\u53d6\u51b3\u4e8e\u6570\u636e\u3002\u4e00\u822c\u60c5\u51b5\u4e0b\uff0c\u5728\u5206\u7c7b\u65f6\u8f83\u5927\u7684K\u503c\u80fd\u591f\u51cf\u5c0f\u566a\u58f0\u7684\u5f71\u54cd\u3002\u4f46\u4f1a\u4f7f\u7c7b\u522b\u4e4b\u95f4\u7684\u754c\u9650\u53d8\u5f97\u6a21\u7cca\u3002\u4e00\u4e2a\u8f83\u597d\u7684K\u503c\u53ef\u901a\u8fc7\u5404\u79cd\u542f\u53d1\u5f0f\u6280\u672f\u6765\u83b7\u53d6\uff0c\u6bd4\u5982\uff0c\u4ea4\u53c9\u9a8c\u8bc1\u3002\u53e6\u5916\u566a\u58f0\u548c\u975e\u76f8\u5173\u6027\u7279\u5f81\u5411\u91cf\u7684\u5b58\u5728\u4f1a\u4f7fK\u8fd1\u90bb\u7b97\u6cd5\u7684\u51c6\u786e\u6027\u51cf\u5c0f\u3002<\/p>\n
<\/p>\n
\u8fd1\u90bb\u7b97\u6cd5\u5177\u6709\u8f83\u5f3a\u7684\u4e00\u81f4\u6027\u7ed3\u679c\u3002\u968f\u7740\u6570\u636e\u8d8b\u4e8e\u65e0\u9650\uff0c\u7b97\u6cd5\u4fdd\u8bc1\u9519\u8bef\u7387\u4e0d\u4f1a\u8d85\u8fc7\u8d1d\u53f6\u65af\u7b97\u6cd5\u9519\u8bef\u7387\u7684\u4e24\u500d\u3002\u5bf9\u4e8e\u4e00\u4e9b\u597d\u7684K\u503c\uff0cK\u8fd1\u90bb\u4fdd\u8bc1\u9519\u8bef\u7387\u4e0d\u4f1a\u8d85\u8fc7\u8d1d\u53f6\u65af\u7406\u8bba\u8bef\u5dee\u7387\u3002<\/p>\n
<\/p>\n
\u6ce8\uff1a\u9a6c\u6c0f\u8ddd\u79bb\u4e00\u5b9a\u8981\u5148\u7ed9\u51fa\u6837\u672c\u96c6\u7684\u7edf\u8ba1\u6027\u8d28\uff0c\u6bd4\u5982\u5747\u503c\u5411\u91cf\uff0c\u534f\u65b9\u5dee\u77e9\u9635\u7b49\u3002\u5173\u4e8e\u9a6c\u6c0f\u8ddd\u79bb\u7684\u4ecb\u7ecd\u5982\u4e0b\uff1a<\/p>\n
\u56fe\u7247<\/p>\n
<\/p>\n
KNN\u7b97\u6cd5\u7684\u4f18\u70b9\uff1a<\/p>\n
1. \u601d\u60f3\u7b80\u5355\uff0c\u7406\u8bba\u6210\u719f\uff0c\u65e2\u53ef\u4ee5\u7528\u6765\u505a\u5206\u7c7b\u4e5f\u53ef\u4ee5\u7528\u6765\u505a\u56de\u5f52\uff1b<\/p>\n
2. \u53ef\u7528\u4e8e\u975e\u7ebf\u6027\u5206\u7c7b\uff1b<\/p>\n
3. \u8bad\u7ec3\u65f6\u95f4\u590d\u6742\u5ea6\u4e3aO(n)\uff1b<\/p>\n
4. \u51c6\u786e\u5ea6\u9ad8\uff0c\u5bf9\u6570\u636e\u6ca1\u6709\u5047\u8bbe\uff0c\u5bf9outlier\u4e0d\u654f\u611f\uff1b<\/p>\n
<\/p>\n
\u7f3a\u70b9\uff1a<\/p>\n
1. \u8ba1\u7b97\u91cf\u5927\uff1b<\/p>\n
2. \u6837\u672c\u4e0d\u5e73\u8861\u95ee\u9898\uff08\u5373\u6709\u4e9b\u7c7b\u522b\u7684\u6837\u672c\u6570\u91cf\u5f88\u591a\uff0c\u800c\u5176\u5b83\u6837\u672c\u7684\u6570\u91cf\u5f88\u5c11\uff09\uff1b<\/p>\n
3. \u9700\u8981\u5927\u91cf\u7684\u5185\u5b58\uff1b<\/p>\n
<\/p>\n
SVM\uff1a<\/p>\n
\u8981\u5b66\u4f1a\u5982\u4f55\u4f7f\u7528libsvm\u4ee5\u53ca\u4e00\u4e9b\u53c2\u6570\u7684\u8c03\u8282\u7ecf\u9a8c\uff0c\u53e6\u5916\u9700\u8981\u7406\u6e05\u695asvm\u7b97\u6cd5\u7684\u4e00\u4e9b\u601d\u8def\uff1a<\/p>\n
<\/p>\n
1. svm\u4e2d\u7684\u6700\u4f18\u5206\u7c7b\u9762\u662f\u5bf9\u6240\u6709\u6837\u672c\u7684\u51e0\u4f55\u88d5\u91cf\u6700\u5927\uff08\u4e3a\u4ec0\u4e48\u8981\u9009\u62e9\u6700\u5927\u95f4\u9694\u5206\u7c7b\u5668\uff0c\u8bf7\u4ece\u6570\u5b66\u89d2\u5ea6\u4e0a\u8bf4\u660e\uff1f\u7f51\u6613\u6df1\u5ea6\u5b66\u4e60\u5c97\u4f4d\u9762\u8bd5\u8fc7\u7a0b\u4e2d\u6709\u88ab\u95ee\u5230\u3002\u7b54\u6848\u5c31\u662f\u51e0\u4f55\u95f4\u9694\u4e0e\u6837\u672c\u7684\u8bef\u5206\u6b21\u6570\u95f4\u5b58\u5728\u5173\u7cfb\uff1a\u56fe\u7247 \uff0c\u5176\u4e2d\u7684\u5206\u6bcd\u5c31\u662f\u6837\u672c\u5230\u5206\u7c7b\u95f4\u9694\u8ddd\u79bb\uff0c\u5206\u5b50\u4e2d\u7684R\u662f\u6240\u6709\u6837\u672c\u4e2d\u7684\u6700\u957f\u5411\u91cf\u503c\uff09\uff0c\u5373\uff1a<\/p>\n
<\/p>\n
\u56fe\u7247<\/p>\n
<\/p>\n
\u7ecf\u8fc7\u4e00\u7cfb\u5217\u63a8\u5bfc\u53ef\u5f97\u4e3a\u4f18\u5316\u4e0b\u9762\u539f\u59cb\u76ee\u6807\uff1a<\/p>\n
\u56fe\u7247<\/p>\n
2. \u4e0b\u9762\u6765\u770b\u770b\u62c9\u683c\u6717\u65e5\u7406\u8bba\uff1a<\/p>\n
\u56fe\u7247<\/p>\n
\u53ef\u4ee5\u5c061\u4e2d\u7684\u4f18\u5316\u76ee\u6807\u8f6c\u6362\u4e3a\u62c9\u683c\u6717\u65e5\u7684\u5f62\u5f0f\uff08\u901a\u8fc7\u5404\u79cd\u5bf9\u5076\u4f18\u5316\uff0cKKD\u6761\u4ef6\uff09\uff0c\u6700\u540e\u76ee\u6807\u51fd\u6570\u4e3a\uff1a<\/p>\n
\u56fe\u7247<\/p>\n
\u6211\u4eec\u53ea\u9700\u8981\u6700\u5c0f\u5316\u4e0a\u8ff0\u76ee\u6807\u51fd\u6570\uff0c\u5176\u4e2d\u7684\u03b1\u4e3a\u539f\u59cb\u4f18\u5316\u95ee\u9898\u4e2d\u7684\u4e0d\u7b49\u5f0f\u7ea6\u675f\u62c9\u683c\u6717\u65e5\u7cfb\u6570\u3002<\/p>\n
<\/p>\n
3. \u5bf92\u4e2d\u6700\u540e\u7684\u5f0f\u5b50\u5206\u522bw\u548cb\u6c42\u5bfc\u53ef\u5f97\uff1a<\/p>\n
\u56fe\u7247<\/p>\n
\u56fe\u7247<\/p>\n
\u7531\u4e0a\u9762\u7b2c1\u5f0f\u5b50\u53ef\u4ee5\u77e5\u9053\uff0c\u5982\u679c\u6211\u4eec\u4f18\u5316\u51fa\u4e86\u03b1\uff0c\u5219\u76f4\u63a5\u53ef\u4ee5\u6c42\u51faw\u4e86\uff0c\u5373\u6a21\u578b\u7684\u53c2\u6570\u641e\u5b9a\u3002\u800c\u4e0a\u9762\u7b2c2\u4e2a\u5f0f\u5b50\u53ef\u4ee5\u4f5c\u4e3a\u540e\u7eed\u4f18\u5316\u7684\u4e00\u4e2a\u7ea6\u675f\u6761\u4ef6\u3002<\/p>\n
<\/p>\n
4. \u5bf92\u4e2d\u6700\u540e\u4e00\u4e2a\u76ee\u6807\u51fd\u6570\u7528\u5bf9\u5076\u4f18\u5316\u7406\u8bba\u53ef\u4ee5\u8f6c\u6362\u4e3a\u4f18\u5316\u4e0b\u9762\u7684\u76ee\u6807\u51fd\u6570\uff1a<\/p>\n
\u56fe\u7247<\/p>\n
\u800c\u8fd9\u4e2a\u51fd\u6570\u53ef\u4ee5\u7528\u5e38\u7528\u7684\u4f18\u5316\u65b9\u6cd5\u6c42\u5f97\u03b1\uff0c\u8fdb\u800c\u6c42\u5f97w\u548cb\u3002<\/p>\n
<\/p>\n
5. \u6309\u7167\u9053\u7406\uff0csvm\u7b80\u5355\u7406\u8bba\u5e94\u8be5\u5230\u6b64\u7ed3\u675f\u3002\u4e0d\u8fc7\u8fd8\u662f\u8981\u8865\u5145\u4e00\u70b9\uff0c\u5373\u5728\u9884\u6d4b\u65f6\u6709\uff1a<\/p>\n
\u56fe\u7247<\/p>\n
\u90a3\u4e2a\u5c16\u62ec\u53f7\u6211\u4eec\u53ef\u4ee5\u7528\u6838\u51fd\u6570\u4ee3\u66ff\uff0c\u8fd9\u4e5f\u662fsvm\u7ecf\u5e38\u548c\u6838\u51fd\u6570\u626f\u5728\u4e00\u8d77\u7684\u539f\u56e0\u3002<\/p>\n
<\/p>\n
6. \u6700\u540e\u662f\u5173\u4e8e\u677e\u5f1b\u53d8\u91cf\u7684\u5f15\u5165\uff0c\u56e0\u6b64\u539f\u59cb\u7684\u76ee\u6807\u4f18\u5316\u516c\u5f0f\u4e3a\uff1a<\/p>\n
\u56fe\u7247<\/p>\n
\u6b64\u65f6\u5bf9\u5e94\u7684\u5bf9\u5076\u4f18\u5316\u516c\u5f0f\u4e3a\uff1a<\/p>\n
\u56fe\u7247<\/p>\n
\u4e0e\u524d\u9762\u7684\u76f8\u6bd4\u53ea\u662f\u03b1\u591a\u4e86\u4e2a\u4e0a\u754c\u3002<\/p>\n
<\/p>\n
SVM\u7b97\u6cd5\u4f18\u70b9\uff1a<\/p>\n
1. \u53ef\u7528\u4e8e\u7ebf\u6027\/\u975e\u7ebf\u6027\u5206\u7c7b\uff0c\u4e5f\u53ef\u4ee5\u7528\u4e8e\u56de\u5f52\uff1b<\/p>\n
2. \u4f4e\u6cdb\u5316\u8bef\u5dee\uff1b<\/p>\n
3. \u5bb9\u6613\u89e3\u91ca\uff1b<\/p>\n
4. \u8ba1\u7b97\u590d\u6742\u5ea6\u8f83\u4f4e\uff1b<\/p>\n
\u7f3a\u70b9\uff1a<\/p>\n
1. \u5bf9\u53c2\u6570\u548c\u6838\u51fd\u6570\u7684\u9009\u62e9\u6bd4\u8f83\u654f\u611f\uff1b<\/p>\n
2. \u539f\u59cb\u7684SVM\u53ea\u6bd4\u8f83\u64c5\u957f\u5904\u7406\u4e8c\u5206\u7c7b\u95ee\u9898\uff1b<\/p>\n
<\/p>\n
Boosting\uff1a<\/p>\n
\u4e3b\u8981\u4ee5Adaboost\u4e3a\u4f8b\uff0c\u9996\u5148\u6765\u770b\u770bAdaboost\u7684\u6d41\u7a0b\u56fe\uff0c\u5982\u4e0b\uff1a<\/p>\n
<\/p>\n
\u56fe\u7247<\/p>\n
<\/p>\n
\u4ece\u56fe\u4e2d\u53ef\u4ee5\u770b\u5230\uff0c\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u6211\u4eec\u9700\u8981\u8bad\u7ec3\u51fa\u591a\u4e2a\u5f31\u5206\u7c7b\u5668\uff08\u56fe\u4e2d\u4e3a3\u4e2a\uff09\uff0c\u6bcf\u4e2a\u5f31\u5206\u7c7b\u5668\u662f\u7531\u4e0d\u540c\u6743\u91cd\u7684\u6837\u672c\uff08\u56fe\u4e2d\u4e3a5\u4e2a\u8bad\u7ec3\u6837\u672c\uff09\u8bad\u7ec3\u5f97\u5230\uff08\u5176\u4e2d\u7b2c\u4e00\u4e2a\u5f31\u5206\u7c7b\u5668\u5bf9\u5e94\u8f93\u5165\u6837\u672c\u7684\u6743\u503c\u662f\u4e00\u6837\u7684\uff09\uff0c\u800c\u6bcf\u4e2a\u5f31\u5206\u7c7b\u5668\u5bf9\u6700\u7ec8\u5206\u7c7b\u7ed3\u679c\u7684\u4f5c\u7528\u4e5f\u4e0d\u540c\uff0c\u662f\u901a\u8fc7\u52a0\u6743\u5e73\u5747\u8f93\u51fa\u7684\uff0c\u6743\u503c\u89c1\u4e0a\u56fe\u4e2d\u4e09\u89d2\u5f62\u91cc\u9762\u7684\u6570\u503c\u3002\u90a3\u4e48\u8fd9\u4e9b\u5f31\u5206\u7c7b\u5668\u548c\u5176\u5bf9\u5e94\u7684\u6743\u503c\u662f\u600e\u6837\u8bad\u7ec3\u51fa\u6765\u7684\u5462\uff1f<\/p>\n
<\/p>\n
\u4e0b\u9762\u901a\u8fc7\u4e00\u4e2a\u4f8b\u5b50\u6765\u7b80\u5355\u8bf4\u660e\uff0c\u5047\u8bbe\u7684\u662f5\u4e2a\u8bad\u7ec3\u6837\u672c\uff0c\u6bcf\u4e2a\u8bad\u7ec3\u6837\u672c\u7684\u7ef4\u5ea6\u4e3a2\uff0c\u5728\u8bad\u7ec3\u7b2c\u4e00\u4e2a\u5206\u7c7b\u5668\u65f65\u4e2a\u6837\u672c\u7684\u6743\u91cd\u5404\u4e3a0.2. \u6ce8\u610f\u8fd9\u91cc\u6837\u672c\u7684\u6743\u503c\u548c\u6700\u7ec8\u8bad\u7ec3\u7684\u5f31\u5206\u7c7b\u5668\u7ec4\u5bf9\u5e94\u7684\u6743\u503c\u03b1\u662f\u4e0d\u540c\u7684\uff0c\u6837\u672c\u7684\u6743\u91cd\u53ea\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u7528\u5230\uff0c\u800c\u03b1\u5728\u8bad\u7ec3\u8fc7\u7a0b\u548c\u6d4b\u8bd5\u8fc7\u7a0b\u90fd\u6709\u7528\u5230\u3002<\/p>\n
<\/p>\n
\u73b0\u5728\u5047\u8bbe\u5f31\u5206\u7c7b\u5668\u662f\u5e26\u4e00\u4e2a\u8282\u70b9\u7684\u7b80\u5355\u51b3\u7b56\u6811\uff0c\u8be5\u51b3\u7b56\u6811\u4f1a\u9009\u62e92\u4e2a\u5c5e\u6027\uff08\u5047\u8bbe\u53ea\u67092\u4e2a\u5c5e\u6027\uff09\u7684\u4e00\u4e2a\uff0c\u7136\u540e\u8ba1\u7b97\u51fa\u8fd9\u4e2a\u5c5e\u6027\u4e2d\u7684\u6700\u4f73\u503c\u7528\u6765\u5206\u7c7b\u3002<\/p>\n
<\/p>\n
Adaboost\u7684\u7b80\u5355\u7248\u672c\u8bad\u7ec3\u8fc7\u7a0b\u5982\u4e0b\uff1a<\/p>\n
<\/p>\n
1. \u8bad\u7ec3\u7b2c\u4e00\u4e2a\u5206\u7c7b\u5668\uff0c\u6837\u672c\u7684\u6743\u503cD\u4e3a\u76f8\u540c\u7684\u5747\u503c\u3002\u901a\u8fc7\u4e00\u4e2a\u5f31\u5206\u7c7b\u5668\uff0c\u5f97\u5230\u8fd95\u4e2a\u6837\u672c\uff08\u8bf7\u5bf9\u5e94\u4e66\u4e2d\u7684\u4f8b\u5b50\u6765\u770b\uff0c\u4f9d\u65e7\u662fmachine learning in action\uff09\u7684\u5206\u7c7b\u9884\u6d4b\u6807\u7b7e\u3002\u4e0e\u7ed9\u51fa\u7684\u6837\u672c\u771f\u5b9e\u6807\u7b7e\u5bf9\u6bd4\uff0c\u5c31\u53ef\u80fd\u51fa\u73b0\u8bef\u5dee(\u5373\u9519\u8bef)\u3002\u5982\u679c\u67d0\u4e2a\u6837\u672c\u9884\u6d4b\u9519\u8bef\uff0c\u5219\u5b83\u5bf9\u5e94\u7684\u9519\u8bef\u503c\u4e3a\u8be5\u6837\u672c\u7684\u6743\u91cd\uff0c\u5982\u679c\u5206\u7c7b\u6b63\u786e\uff0c\u5219\u9519\u8bef\u503c\u4e3a0. \u6700\u540e\u7d2f\u52a05\u4e2a\u6837\u672c\u7684\u9519\u8bef\u7387\u4e4b\u548c\uff0c\u8bb0\u4e3a\u03b5\u3002<\/p>\n
<\/p>\n
2. \u901a\u8fc7\u03b5\u6765\u8ba1\u7b97\u8be5\u5f31\u5206\u7c7b\u5668\u7684\u6743\u91cd\u03b1\uff0c\u516c\u5f0f\u5982\u4e0b\uff1a<\/p>\n
\u56fe\u7247<\/p>\n
3. \u901a\u8fc7\u03b1\u6765\u8ba1\u7b97\u8bad\u7ec3\u4e0b\u4e00\u4e2a\u5f31\u5206\u7c7b\u5668\u6837\u672c\u7684\u6743\u91cdD\uff0c\u5982\u679c\u5bf9\u5e94\u6837\u672c\u5206\u7c7b\u6b63\u786e\uff0c\u5219\u51cf\u5c0f\u8be5\u6837\u672c\u7684\u6743\u91cd\uff0c\u516c\u5f0f\u4e3a\uff1a<\/p>\n
\u56fe\u7247<\/p>\n
\u5982\u679c\u6837\u672c\u5206\u7c7b\u9519\u8bef\uff0c\u5219\u589e\u52a0\u8be5\u6837\u672c\u7684\u6743\u91cd\uff0c\u516c\u5f0f\u4e3a\uff1a<\/p>\n
\u56fe\u7247<\/p>\n
4. \u5faa\u73af\u6b65\u9aa41,2,3\u6765\u7ee7\u7eed\u8bad\u7ec3\u591a\u4e2a\u5206\u7c7b\u5668\uff0c\u53ea\u662f\u5176D\u503c\u4e0d\u540c\u800c\u5df2\u3002<\/p>\n
\u6d4b\u8bd5\u8fc7\u7a0b\u5982\u4e0b\uff1a<\/p>\n
\u8f93\u5165\u4e00\u4e2a\u6837\u672c\u5230\u8bad\u7ec3\u597d\u7684\u6bcf\u4e2a\u5f31\u5206\u7c7b\u4e2d\uff0c\u5219\u6bcf\u4e2a\u5f31\u5206\u7c7b\u90fd\u5bf9\u5e94\u4e00\u4e2a\u8f93\u51fa\u6807\u7b7e\uff0c\u7136\u540e\u8be5\u6807\u7b7e\u4e58\u4ee5\u5bf9\u5e94\u7684\u03b1\uff0c\u6700\u540e\u6c42\u548c\u5f97\u5230\u503c\u7684\u7b26\u53f7\u5373\u4e3a\u9884\u6d4b\u6807\u7b7e\u503c\u3002<\/p>\n
<\/p>\n
Boosting\u7b97\u6cd5\u7684\u4f18\u70b9\uff1a<\/p>\n
1. \u4f4e\u6cdb\u5316\u8bef\u5dee\uff1b<\/p>\n
2. \u5bb9\u6613\u5b9e\u73b0\uff0c\u5206\u7c7b\u51c6\u786e\u7387\u8f83\u9ad8\uff0c\u6ca1\u6709\u592a\u591a\u53c2\u6570\u53ef\u4ee5\u8c03\uff1b<\/p>\n
3. \u7f3a\u70b9\uff1a<\/p>\n
4. \u5bf9outlier\u6bd4\u8f83\u654f\u611f\uff1b<\/p>\n
<\/p>\n
\u805a\u7c7b\uff1a<\/p>\n
\u6839\u636e\u805a\u7c7b\u601d\u60f3\u5212\u5206\uff1a<\/p>\n
1. \u57fa\u4e8e\u5212\u5206\u7684\u805a\u7c7b:<\/p>\n
K-means, k-medoids(\u6bcf\u4e00\u4e2a\u7c7b\u522b\u4e2d\u627e\u4e00\u4e2a\u6837\u672c\u70b9\u6765\u4ee3\u8868),CLARANS.<\/p>\n
k-means\u662f\u4f7f\u4e0b\u9762\u7684\u8868\u8fbe\u5f0f\u503c\u6700\u5c0f\uff1a<\/p>\n
\u56fe\u7247<\/p>\n
k-means\u7b97\u6cd5\u7684\u4f18\u70b9\uff1a<\/p>\n
\uff081\uff09k-means\u7b97\u6cd5\u662f\u89e3\u51b3\u805a\u7c7b\u95ee\u9898\u7684\u4e00\u79cd\u7ecf\u5178\u7b97\u6cd5\uff0c\u7b97\u6cd5\u7b80\u5355\u3001\u5feb\u901f\u3002<\/p>\n
\uff082\uff09\u5bf9\u5904\u7406\u5927\u6570\u636e\u96c6\uff0c\u8be5\u7b97\u6cd5\u662f\u76f8\u5bf9\u53ef\u4f38\u7f29\u7684\u548c\u9ad8\u6548\u7387\u7684\uff0c\u56e0\u4e3a\u5b83\u7684\u590d\u6742\u5ea6\u5927\u7ea6\u662fO(nkt)\uff0c\u5176\u4e2dn\u662f\u6240\u6709\u5bf9\u8c61\u7684\u6570\u76ee\uff0ck\u662f\u7c07\u7684\u6570\u76ee,t\u662f\u8fed\u4ee3\u7684\u6b21\u6570\u3002\u901a\u5e38k<<n\u3002\u8fd9\u4e2a\u7b97\u6cd5\u901a\u5e38\u5c40\u90e8\u6536\u655b\u3002<\/p>\n
\uff083\uff09\u7b97\u6cd5\u5c1d\u8bd5\u627e\u51fa\u4f7f\u5e73\u65b9\u8bef\u5dee\u51fd\u6570\u503c\u6700\u5c0f\u7684k\u4e2a\u5212\u5206\u3002\u5f53\u7c07\u662f\u5bc6\u96c6\u7684\u3001\u7403\u72b6\u6216\u56e2\u72b6\u7684\uff0c\u4e14\u7c07\u4e0e\u7c07\u4e4b\u95f4\u533a\u522b\u660e\u663e\u65f6\uff0c\u805a\u7c7b\u6548\u679c\u8f83\u597d\u3002<\/p>\n
<\/p>\n
\u7f3a\u70b9\uff1a<\/p>\n
\uff081\uff09k-\u5e73\u5747\u65b9\u6cd5\u53ea\u6709\u5728\u7c07\u7684\u5e73\u5747\u503c\u88ab\u5b9a\u4e49\u7684\u60c5\u51b5\u4e0b\u624d\u80fd\u4f7f\u7528\uff0c\u4e14\u5bf9\u6709\u4e9b\u5206\u7c7b\u5c5e\u6027\u7684\u6570\u636e\u4e0d\u9002\u5408\u3002<\/p>\n
\uff082\uff09\u8981\u6c42\u7528\u6237\u5fc5\u987b\u4e8b\u5148\u7ed9\u51fa\u8981\u751f\u6210\u7684\u7c07\u7684\u6570\u76eek\u3002<\/p>\n
\uff083\uff09\u5bf9\u521d\u503c\u654f\u611f\uff0c\u5bf9\u4e8e\u4e0d\u540c\u7684\u521d\u59cb\u503c\uff0c\u53ef\u80fd\u4f1a\u5bfc\u81f4\u4e0d\u540c\u7684\u805a\u7c7b\u7ed3\u679c\u3002<\/p>\n
\uff084\uff09\u4e0d\u9002\u5408\u4e8e\u53d1\u73b0\u975e\u51f8\u9762\u5f62\u72b6\u7684\u7c07\uff0c\u6216\u8005\u5927\u5c0f\u5dee\u522b\u5f88\u5927\u7684\u7c07\u3002<\/p>\n
\uff085\uff09\u5bf9\u4e8e”\u566a\u58f0”\u548c\u5b64\u7acb\u70b9\u6570\u636e\u654f\u611f\uff0c\u5c11\u91cf\u7684\u8be5\u7c7b\u6570\u636e\u80fd\u591f\u5bf9\u5e73\u5747\u503c\u4ea7\u751f\u6781\u5927\u5f71\u54cd\u3002<\/p>\n
<\/p>\n
2. \u57fa\u4e8e\u5c42\u6b21\u7684\u805a\u7c7b\uff1a<\/p>\n
\u81ea\u5e95\u5411\u4e0a\u7684\u51dd\u805a\u65b9\u6cd5\uff0c\u6bd4\u5982AGNES\u3002<\/p>\n
\u81ea\u4e0a\u5411\u4e0b\u7684\u5206\u88c2\u65b9\u6cd5\uff0c\u6bd4\u5982DIANA\u3002<\/p>\n
<\/p>\n
3. \u57fa\u4e8e\u5bc6\u5ea6\u7684\u805a\u7c7b\uff1aDBSACN,OPTICS,BIRCH(CF-Tree),CURE.<\/p>\n
4. \u57fa\u4e8e\u7f51\u683c\u7684\u65b9\u6cd5\uff1aSTING, WaveCluster.<\/p>\n
5. \u57fa\u4e8e\u6a21\u578b\u7684\u805a\u7c7b\uff1aEM,SOM,COBWEB.<\/p>\n
<\/p>\n
\u63a8\u8350\u7cfb\u7edf\uff1a\u63a8\u8350\u7cfb\u7edf\u7684\u5b9e\u73b0\u4e3b\u8981\u5206\u4e3a\u4e24\u4e2a\u65b9\u9762\uff1a\u57fa\u4e8e\u5185\u5bb9\u7684\u5b9e\u73b0\u548c\u534f\u540c\u6ee4\u6ce2\u7684\u5b9e\u73b0\u3002<\/p>\n
<\/p>\n
\u57fa\u4e8e\u5185\u5bb9\u7684\u5b9e\u73b0\uff1a\u4e0d\u540c\u4eba\u5bf9\u4e0d\u540c\u7535\u5f71\u7684\u8bc4\u5206\u8fd9\u4e2a\u4f8b\u5b50\uff0c\u53ef\u4ee5\u770b\u505a\u662f\u4e00\u4e2a\u666e\u901a\u7684\u56de\u5f52\u95ee\u9898\uff0c\u56e0\u6b64\u6bcf\u90e8\u7535\u5f71\u90fd\u9700\u8981\u63d0\u524d\u63d0\u53d6\u51fa\u4e00\u4e2a\u7279\u5f81\u5411\u91cf(\u5373x\u503c)\uff0c\u7136\u540e\u9488\u5bf9\u6bcf\u4e2a\u7528\u6237\u5efa\u6a21\uff0c\u5373\u6bcf\u4e2a\u7528\u6237\u6253\u7684\u5206\u503c\u4f5c\u4e3ay\u503c\uff0c\u5229\u7528\u8fd9\u4e9b\u5df2\u6709\u7684\u5206\u503cy\u548c\u7535\u5f71\u7279\u5f81\u503cx\u5c31\u53ef\u4ee5\u8bad\u7ec3\u56de\u5f52\u6a21\u578b\u4e86(\u6700\u5e38\u89c1\u7684\u5c31\u662f\u7ebf\u6027\u56de\u5f52)\u3002<\/p>\n
<\/p>\n
\u8fd9\u6837\u5c31\u53ef\u4ee5\u9884\u6d4b\u90a3\u4e9b\u7528\u6237\u6ca1\u6709\u8bc4\u5206\u7684\u7535\u5f71\u7684\u5206\u6570\u3002\uff08\u503c\u5f97\u6ce8\u610f\u7684\u662f\u9700\u5bf9\u6bcf\u4e2a\u7528\u6237\u90fd\u5efa\u7acb\u4ed6\u81ea\u5df1\u7684\u56de\u5f52\u6a21\u578b\uff09<\/p>\n
<\/p>\n
\u4ece\u53e6\u4e00\u4e2a\u89d2\u5ea6\u6765\u770b\uff0c\u4e5f\u53ef\u4ee5\u662f\u5148\u7ed9\u5b9a\u6bcf\u4e2a\u7528\u6237\u5bf9\u67d0\u79cd\u7535\u5f71\u7684\u559c\u597d\u7a0b\u5ea6\uff08\u5373\u6743\u503c\uff09\uff0c\u7136\u540e\u5b66\u51fa\u6bcf\u90e8\u7535\u5f71\u7684\u7279\u5f81\uff0c\u6700\u540e\u91c7\u7528\u56de\u5f52\u6765\u9884\u6d4b\u90a3\u4e9b\u6ca1\u6709\u88ab\u8bc4\u5206\u7684\u7535\u5f71\u3002<\/p>\n
<\/p>\n
\u5f53\u7136\u8fd8\u53ef\u4ee5\u662f\u540c\u65f6\u4f18\u5316\u5f97\u5230\u6bcf\u4e2a\u7528\u6237\u5bf9\u4e0d\u540c\u7c7b\u578b\u7535\u5f71\u7684\u70ed\u7231\u7a0b\u5ea6\u4ee5\u53ca\u6bcf\u90e8\u7535\u5f71\u7684\u7279\u5f81\u3002<\/p>\n
<\/p>\n
\u57fa\u4e8e\u534f\u540c\u6ee4\u6ce2\u7684\u5b9e\u73b0\uff1a\u534f\u540c\u6ee4\u6ce2\uff08CF\uff09\u53ef\u4ee5\u770b\u505a\u662f\u4e00\u4e2a\u5206\u7c7b\u95ee\u9898\uff0c\u4e5f\u53ef\u4ee5\u770b\u505a\u662f\u77e9\u9635\u5206\u89e3\u95ee\u9898\u3002\u534f\u540c\u6ee4\u6ce2\u4e3b\u8981\u662f\u57fa\u4e8e\u6bcf\u4e2a\u4eba\u81ea\u5df1\u7684\u559c\u597d\u90fd\u7c7b\u4f3c\u8fd9\u4e00\u7279\u5f81\uff0c\u5b83\u4e0d\u4f9d\u8d56\u4e8e\u4e2a\u4eba\u7684\u57fa\u672c\u4fe1\u606f\u3002<\/p>\n
<\/p>\n
\u6bd4\u5982\u521a\u521a\u90a3\u4e2a\u7535\u5f71\u8bc4\u5206\u7684\u4f8b\u5b50\u4e2d\uff0c\u9884\u6d4b\u90a3\u4e9b\u6ca1\u6709\u88ab\u8bc4\u5206\u7684\u7535\u5f71\u7684\u5206\u6570\u53ea\u4f9d\u8d56\u4e8e\u5df2\u7ecf\u6253\u5206\u7684\u90a3\u4e9b\u5206\u6570\uff0c\u5e76\u4e0d\u9700\u8981\u53bb\u5b66\u4e60\u90a3\u4e9b\u7535\u5f71\u7684\u7279\u5f81\u3002<\/p>\n
<\/p>\n
SVD\u5c06\u77e9\u9635\u5206\u89e3\u4e3a\u4e09\u4e2a\u77e9\u9635\u7684\u4e58\u79ef\uff0c\u516c\u5f0f\u5982\u4e0b\u6240\u793a\uff1a<\/p>\n
<\/p>\n
\u56fe\u7247<\/p>\n
<\/p>\n
\u4e2d\u95f4\u7684\u77e9\u9635sigma\u4e3a\u5bf9\u89d2\u77e9\u9635\uff0c\u5bf9\u89d2\u5143\u7d20\u7684\u503c\u4e3aData\u77e9\u9635\u7684\u5947\u5f02\u503c(\u6ce8\u610f\u5947\u5f02\u503c\u548c\u7279\u5f81\u503c\u662f\u4e0d\u540c\u7684)\uff0c\u4e14\u5df2\u7ecf\u4ece\u5927\u5230\u5c0f\u6392\u5217\u597d\u4e86\u3002\u5373\u4f7f\u53bb\u6389\u7279\u5f81\u503c\u5c0f\u7684\u90a3\u4e9b\u7279\u5f81\uff0c\u4f9d\u7136\u53ef\u4ee5\u5f88\u597d\u7684\u91cd\u6784\u51fa\u539f\u59cb\u77e9\u9635\u3002\u5982\u4e0b\u56fe\u6240\u793a\uff1a<\/p>\n
<\/p>\n
\u56fe\u7247<\/p>\n
<\/p>\n
\u5176\u4e2d\u66f4\u6df1\u7684\u989c\u8272\u4ee3\u8868\u53bb\u6389\u5c0f\u7279\u5f81\u503c\u91cd\u6784\u65f6\u7684\u4e09\u4e2a\u77e9\u9635\u3002<\/p>\n
<\/p>\n
\u679cm\u4ee3\u8868\u5546\u54c1\u7684\u4e2a\u6570\uff0cn\u4ee3\u8868\u7528\u6237\u7684\u4e2a\u6570\uff0c\u5219U\u77e9\u9635\u7684\u6bcf\u4e00\u884c\u4ee3\u8868\u5546\u54c1\u7684\u5c5e\u6027\uff0c\u73b0\u5728\u901a\u8fc7\u964d\u7ef4U\u77e9\u9635\uff08\u53d6\u6df1\u8272\u90e8\u5206\uff09\u540e\uff0c\u6bcf\u4e00\u4e2a\u5546\u54c1\u7684\u5c5e\u6027\u53ef\u4ee5\u7528\u66f4\u4f4e\u7684\u7ef4\u5ea6\u8868\u793a\uff08\u5047\u8bbe\u4e3ak\u7ef4\uff09\u3002\u8fd9\u6837\u5f53\u65b0\u6765\u4e00\u4e2a\u7528\u6237\u7684\u5546\u54c1\u63a8\u8350\u5411\u91cfX\uff0c\u5219\u53ef\u4ee5\u6839\u636e\u516c\u5f0fX’*U1*inv(S1)\u5f97\u5230\u4e00\u4e2ak\u7ef4\u7684\u5411\u91cf\uff0c\u7136\u540e\u5728V\u2019\u4e2d\u5bfb\u627e\u6700\u76f8\u4f3c\u7684\u90a3\u4e00\u4e2a\u7528\u6237\uff08\u76f8\u4f3c\u5ea6\u6d4b\u91cf\u53ef\u7528\u4f59\u5f26\u516c\u5f0f\u7b49\uff09\uff0c\u6839\u636e\u8fd9\u4e2a\u7528\u6237\u7684\u8bc4\u5206\u6765\u63a8\u8350\uff08\u4e3b\u8981\u662f\u63a8\u8350\u65b0\u7528\u6237\u672a\u6253\u5206\u7684\u90a3\u4e9b\u5546\u54c1\uff09\u3002<\/p>\n
<\/p>\n
pLSA\uff1a\u7531LSA\u53d1\u5c55\u8fc7\u6765\uff0c\u800c\u65e9\u671fLSA\u7684\u5b9e\u73b0\u4e3b\u8981\u662f\u901a\u8fc7SVD\u5206\u89e3\u3002pLSA\u7684\u6a21\u578b\u56fe\u5982\u4e0b<\/p>\n
<\/p>\n
\u56fe\u7247<\/p>\n
\u516c\u5f0f\u4e2d\u7684\u610f\u4e49\u5982\u4e0b\uff1a<\/p>\n
\u56fe\u7247<\/p>\n
<\/p>\n
LDA\u4e3b\u9898\u6a21\u578b\uff0c\u6982\u7387\u56fe\u5982\u4e0b\uff1a<\/p>\n
<\/p>\n
\u56fe\u7247<\/p>\n
<\/p>\n
\u548cpLSA\u4e0d\u540c\u7684\u662fLDA\u4e2d\u5047\u8bbe\u4e86\u5f88\u591a\u5148\u9a8c\u5206\u5e03\uff0c\u4e14\u4e00\u822c\u53c2\u6570\u7684\u5148\u9a8c\u5206\u5e03\u90fd\u5047\u8bbe\u4e3aDirichlet\u5206\u5e03\uff0c\u5176\u539f\u56e0\u662f\u5171\u8f6d\u5206\u5e03\u65f6\u5148\u9a8c\u6982\u7387\u548c\u540e\u9a8c\u6982\u7387\u7684\u5f62\u5f0f\u76f8\u540c\u3002<\/p>\n
<\/p>\n
GDBT\uff1aGBDT(Gradient Boosting Decision Tree) \u53c8\u53eb MART\uff08Multiple Additive Regression Tree)\uff0c\u597d\u50cf\u5728\u963f\u91cc\u5185\u90e8\u7528\u5f97\u6bd4\u8f83\u591a\uff08\u6240\u4ee5\u963f\u91cc\u7b97\u6cd5\u5c97\u4f4d\u9762\u8bd5\u65f6\u53ef\u80fd\u4f1a\u95ee\u5230\uff09\uff0c\u5b83\u662f\u4e00\u79cd\u8fed\u4ee3\u7684\u51b3\u7b56\u6811\u7b97\u6cd5\uff0c\u8be5\u7b97\u6cd5\u7531\u591a\u68f5\u51b3\u7b56\u6811\u7ec4\u6210\uff0c\u6240\u6709\u6811\u7684\u8f93\u51fa\u7ed3\u679c\u7d2f\u52a0\u8d77\u6765\u5c31\u662f\u6700\u7ec8\u7b54\u6848\u3002<\/p>\n
<\/p>\n
\u5b83\u5728\u88ab\u63d0\u51fa\u4e4b\u521d\u5c31\u548cSVM\u4e00\u8d77\u88ab\u8ba4\u4e3a\u662f\u6cdb\u5316\u80fd\u529b\uff08generalization)\u8f83\u5f3a\u7684\u7b97\u6cd5\u3002\u8fd1\u4e9b\u5e74\u66f4\u56e0\u4e3a\u88ab\u7528\u4e8e\u641c\u7d22\u6392\u5e8f\u7684\u673a\u5668\u5b66\u4e60\u6a21\u578b\u800c\u5f15\u8d77\u5927\u5bb6\u5173\u6ce8\u3002<\/p>\n
<\/p>\n
GBDT\u662f\u56de\u5f52\u6811\uff0c\u4e0d\u662f\u5206\u7c7b\u6811\u3002\u5176\u6838\u5fc3\u5c31\u5728\u4e8e\uff0c\u6bcf\u4e00\u68f5\u6811\u662f\u4ece\u4e4b\u524d\u6240\u6709\u6811\u7684\u6b8b\u5dee\u4e2d\u6765\u5b66\u4e60\u7684\u3002\u4e3a\u4e86\u9632\u6b62\u8fc7\u62df\u5408\uff0c\u548cAdaboosting\u4e00\u6837\uff0c\u4e5f\u52a0\u5165\u4e86boosting\u8fd9\u4e00\u9879\u3002<\/p>\n
<\/p>\n
<\/p>\n
Regularization\u4f5c\u7528\u662f<\/p>\n
1. \u6570\u503c\u4e0a\u66f4\u5bb9\u6613\u6c42\u89e3\uff1b<\/p>\n
2. \u7279\u5f81\u6570\u76ee\u592a\u5927\u65f6\u66f4\u7a33\u5b9a\uff1b<\/p>\n
3. \u63a7\u5236\u6a21\u578b\u7684\u590d\u6742\u5ea6\uff0c\u5149\u6ed1\u6027\u3002\u590d\u6742\u6027\u8d8a\u5c0f\u4e14\u8d8a\u5149\u6ed1\u7684\u76ee\u6807\u51fd\u6570\u6cdb\u5316\u80fd\u529b\u8d8a\u5f3a\u3002\u800c\u52a0\u5165\u89c4\u5219\u9879\u80fd\u4f7f\u76ee\u6807\u51fd\u6570\u590d\u6742\u5ea6\u51cf\u5c0f\uff0c\u4e14\u66f4\u5149\u6ed1\u3002<\/p>\n
4. \u51cf\u5c0f\u53c2\u6570\u7a7a\u95f4\uff1b\u53c2\u6570\u7a7a\u95f4\u8d8a\u5c0f\uff0c\u590d\u6742\u5ea6\u8d8a\u4f4e\u3002<\/p>\n
5. \u7cfb\u6570\u8d8a\u5c0f\uff0c\u6a21\u578b\u8d8a\u7b80\u5355\uff0c\u800c\u6a21\u578b\u8d8a\u7b80\u5355\u5219\u6cdb\u5316\u80fd\u529b\u8d8a\u5f3a\uff08Ng\u5b8f\u89c2\u4e0a\u7ed9\u51fa\u7684\u89e3\u91ca\uff09\u3002<\/p>\n
6. \u53ef\u4ee5\u770b\u6210\u662f\u6743\u503c\u7684\u9ad8\u65af\u5148\u9a8c\u3002<\/p>\n
<\/p>\n
\u5f02\u5e38\u68c0\u6d4b\uff1a\u53ef\u4ee5\u4f30\u8ba1\u6837\u672c\u7684\u5bc6\u5ea6\u51fd\u6570\uff0c\u5bf9\u4e8e\u65b0\u6837\u672c\u76f4\u63a5\u8ba1\u7b97\u5176\u5bc6\u5ea6\uff0c\u5982\u679c\u5bc6\u5ea6\u503c\u5c0f\u4e8e\u67d0\u4e00\u9608\u503c\uff0c\u5219\u8868\u793a\u8be5\u6837\u672c\u5f02\u5e38\u3002\u800c\u5bc6\u5ea6\u51fd\u6570\u4e00\u822c\u91c7\u7528\u591a\u7ef4\u7684\u9ad8\u65af\u5206\u5e03\u3002<\/p>\n
<\/p>\n
\u5982\u679c\u6837\u672c\u6709n\u7ef4\uff0c\u5219\u6bcf\u4e00\u7ef4\u7684\u7279\u5f81\u90fd\u53ef\u4ee5\u770b\u4f5c\u662f\u7b26\u5408\u9ad8\u65af\u5206\u5e03\u7684\uff0c\u5373\u4f7f\u8fd9\u4e9b\u7279\u5f81\u53ef\u89c6\u5316\u51fa\u6765\u4e0d\u592a\u7b26\u5408\u9ad8\u65af\u5206\u5e03\uff0c\u4e5f\u53ef\u4ee5\u5bf9\u8be5\u7279\u5f81\u8fdb\u884c\u6570\u5b66\u8f6c\u6362\u8ba9\u5176\u770b\u8d77\u6765\u50cf\u9ad8\u65af\u5206\u5e03\uff0c\u6bd4\u5982\u8bf4x=log(x+c), x=x^(1\/c)\u7b49\u3002\u5f02\u5e38\u68c0\u6d4b\u7684\u7b97\u6cd5\u6d41\u7a0b\u5982\u4e0b\uff1a<\/p>\n
<\/p>\n
\u56fe\u7247<\/p>\n
<\/p>\n
\u5176\u4e2d\u7684\u03b5\u4e5f\u662f\u901a\u8fc7\u4ea4\u53c9\u9a8c\u8bc1\u5f97\u5230\u7684\uff0c\u4e5f\u5c31\u662f\u8bf4\u5728\u8fdb\u884c\u5f02\u5e38\u68c0\u6d4b\u65f6\uff0c\u524d\u9762\u7684p(x)\u7684\u5b66\u4e60\u662f\u7528\u7684\u65e0\u76d1\u7763\uff0c\u540e\u9762\u7684\u53c2\u6570\u03b5\u5b66\u4e60\u662f\u7528\u7684\u6709\u76d1\u7763\u3002\u90a3\u4e48\u4e3a\u4ec0\u4e48\u4e0d\u5168\u90e8\u4f7f\u7528\u666e\u901a\u6709\u76d1\u7763\u7684\u65b9\u6cd5\u6765\u5b66\u4e60\u5462\uff08\u5373\u628a\u5b83\u770b\u505a\u662f\u4e00\u4e2a\u666e\u901a\u7684\u4e8c\u5206\u7c7b\u95ee\u9898\uff09\uff1f<\/p>\n
<\/p>\n
\u4e3b\u8981\u662f\u56e0\u4e3a\u5728\u5f02\u5e38\u68c0\u6d4b\u4e2d\uff0c\u5f02\u5e38\u7684\u6837\u672c\u6570\u91cf\u975e\u5e38\u5c11\u800c\u6b63\u5e38\u6837\u672c\u6570\u91cf\u975e\u5e38\u591a\uff0c\u56e0\u6b64\u4e0d\u8db3\u4ee5\u5b66\u4e60\u5230\u597d\u7684\u5f02\u5e38\u884c\u4e3a\u6a21\u578b\u7684\u53c2\u6570\uff0c\u56e0\u4e3a\u540e\u9762\u65b0\u6765\u7684\u5f02\u5e38\u6837\u672c\u53ef\u80fd\u5b8c\u5168\u662f\u4e0e\u8bad\u7ec3\u6837\u672c\u4e2d\u7684\u6a21\u5f0f\u4e0d\u540c\u3002<\/p>\n
<\/p>\n
<\/p>\n
EM\u7b97\u6cd5\uff1a\u6709\u65f6\u5019\u56e0\u4e3a\u6837\u672c\u7684\u4ea7\u751f\u548c\u9690\u542b\u53d8\u91cf\u6709\u5173\uff08\u9690\u542b\u53d8\u91cf\u662f\u4e0d\u80fd\u89c2\u5bdf\u7684\uff09\uff0c\u800c\u6c42\u6a21\u578b\u7684\u53c2\u6570\u65f6\u4e00\u822c\u91c7\u7528\u6700\u5927\u4f3c\u7136\u4f30\u8ba1\uff0c\u7531\u4e8e\u542b\u6709\u4e86\u9690\u542b\u53d8\u91cf\uff0c\u6240\u4ee5\u5bf9\u4f3c\u7136\u51fd\u6570\u53c2\u6570\u6c42\u5bfc\u662f\u6c42\u4e0d\u51fa\u6765\u7684\uff0c\u8fd9\u65f6\u53ef\u4ee5\u91c7\u7528EM\u7b97\u6cd5\u6765\u6c42\u6a21\u578b\u7684\u53c2\u6570\u7684\uff08\u5bf9\u5e94\u6a21\u578b\u53c2\u6570\u4e2a\u6570\u53ef\u80fd\u6709\u591a\u4e2a\uff09\uff0c<\/p>\n
<\/p>\n
EM\u7b97\u6cd5\u4e00\u822c\u5206\u4e3a2\u6b65\uff1a<\/p>\n
E\u6b65\uff1a\u9009\u53d6\u4e00\u7ec4\u53c2\u6570\uff0c\u6c42\u51fa\u5728\u8be5\u53c2\u6570\u4e0b\u9690\u542b\u53d8\u91cf\u7684\u6761\u4ef6\u6982\u7387\u503c\uff1b<\/p>\n
M\u6b65\uff1a\u7ed3\u5408E\u6b65\u6c42\u51fa\u7684\u9690\u542b\u53d8\u91cf\u6761\u4ef6\u6982\u7387\uff0c\u6c42\u51fa\u4f3c\u7136\u51fd\u6570\u4e0b\u754c\u51fd\u6570\uff08\u672c\u8d28\u4e0a\u662f\u67d0\u4e2a\u671f\u671b\u51fd\u6570\uff09\u7684\u6700\u5927\u503c\u3002<\/p>\n
<\/p>\n
\u91cd\u590d\u4e0a\u97622\u6b65\u76f4\u81f3\u6536\u655b\uff0c\u516c\u5f0f\u5982\u4e0b\u6240\u793a\uff1a<\/p>\n
<\/p>\n
\u56fe\u7247<\/p>\n
<\/p>\n
M\u6b65\u516c\u5f0f\u4e2d\u4e0b\u754c\u51fd\u6570\u7684\u63a8\u5bfc\u8fc7\u7a0b\uff1a<\/p>\n
<\/p>\n
\u56fe\u7247<\/p>\n
<\/p>\n
EM\u7b97\u6cd5\u4e00\u4e2a\u5e38\u89c1\u7684\u4f8b\u5b50\u5c31\u662fGMM\u6a21\u578b\uff0c\u6bcf\u4e2a\u6837\u672c\u90fd\u6709\u53ef\u80fd\u7531k\u4e2a\u9ad8\u65af\u4ea7\u751f\uff0c\u53ea\u4e0d\u8fc7\u7531\u6bcf\u4e2a\u9ad8\u65af\u4ea7\u751f\u7684\u6982\u7387\u4e0d\u540c\u800c\u5df2\uff0c\u56e0\u6b64\u6bcf\u4e2a\u6837\u672c\u90fd\u6709\u5bf9\u5e94\u7684\u9ad8\u65af\u5206\u5e03\uff08k\u4e2a\u4e2d\u7684\u67d0\u4e00\u4e2a\uff09\uff0c\u6b64\u65f6\u7684\u9690\u542b\u53d8\u91cf\u5c31\u662f\u6bcf\u4e2a\u6837\u672c\u5bf9\u5e94\u7684\u67d0\u4e2a\u9ad8\u65af\u5206\u5e03\u3002<\/p>\n
<\/p>\n
GMM\u7684E\u6b65\u516c\u5f0f\u5982\u4e0b\uff08\u8ba1\u7b97\u6bcf\u4e2a\u6837\u672c\u5bf9\u5e94\u6bcf\u4e2a\u9ad8\u65af\u7684\u6982\u7387\uff09\uff1a<\/p>\n
<\/p>\n
\u56fe\u7247<\/p>\n
<\/p>\n
\u66f4\u5177\u4f53\u7684\u8ba1\u7b97\u516c\u5f0f\u4e3a\uff1a<\/p>\n
\u56fe\u7247<\/p>\n
<\/p>\n
M\u6b65\u516c\u5f0f\u5982\u4e0b\uff08\u8ba1\u7b97\u6bcf\u4e2a\u9ad8\u65af\u7684\u6bd4\u91cd\uff0c\u5747\u503c\uff0c\u65b9\u5dee\u8fd93\u4e2a\u53c2\u6570\uff09\uff1a<\/p>\n
\u56fe\u7247<\/p>\n
<\/p>\n
<\/p>\n
Apriori\u662f\u5173\u8054\u5206\u6790\u4e2d\u6bd4\u8f83\u65e9\u7684\u4e00\u79cd\u65b9\u6cd5\uff0c\u4e3b\u8981\u7528\u6765\u6316\u6398\u90a3\u4e9b\u9891\u7e41\u9879\u96c6\u5408\u3002\u5176\u601d\u60f3\u662f\uff1a<\/p>\n
1. \u5982\u679c\u4e00\u4e2a\u9879\u76ee\u96c6\u5408\u4e0d\u662f\u9891\u7e41\u96c6\u5408\uff0c\u90a3\u4e48\u4efb\u4f55\u5305\u542b\u5b83\u7684\u9879\u76ee\u96c6\u5408\u4e5f\u4e00\u5b9a\u4e0d\u662f\u9891\u7e41\u96c6\u5408\uff1b<\/p>\n
2. \u5982\u679c\u4e00\u4e2a\u9879\u76ee\u96c6\u5408\u662f\u9891\u7e41\u96c6\u5408\uff0c\u90a3\u4e48\u5b83\u7684\u4efb\u4f55\u975e\u7a7a\u5b50\u96c6\u4e5f\u662f\u9891\u7e41\u96c6\u5408\uff1b<\/p>\n
<\/p>\n
Aprioir\u9700\u8981\u626b\u63cf\u9879\u76ee\u8868\u591a\u904d\uff0c\u4ece\u4e00\u4e2a\u9879\u76ee\u5f00\u59cb\u626b\u63cf\uff0c\u820d\u53bb\u6389\u90a3\u4e9b\u4e0d\u662f\u9891\u7e41\u7684\u9879\u76ee\uff0c\u5f97\u5230\u7684\u96c6\u5408\u79f0\u4e3aL\uff0c\u7136\u540e\u5bf9L\u4e2d\u7684\u6bcf\u4e2a\u5143\u7d20\u8fdb\u884c\u81ea\u7ec4\u5408\uff0c\u751f\u6210\u6bd4\u4e0a\u6b21\u626b\u63cf\u591a\u4e00\u4e2a\u9879\u76ee\u7684\u96c6\u5408\uff0c\u8be5\u96c6\u5408\u79f0\u4e3aC\uff0c\u63a5\u7740\u53c8\u626b\u63cf\u53bb\u6389\u90a3\u4e9b\u975e\u9891\u7e41\u7684\u9879\u76ee\uff0c\u91cd\u590d\u2026<\/p>\n
<\/p>\n
\u770b\u4e0b\u9762\u8fd9\u4e2a\u4f8b\u5b50\uff0c\u5143\u7d20\u9879\u76ee\u8868\u683c\uff1a<\/p>\n
\u56fe\u7247<\/p>\n
<\/p>\n
\u5982\u679c\u6bcf\u4e2a\u6b65\u9aa4\u4e0d\u53bb\u6389\u975e\u9891\u7e41\u9879\u76ee\u96c6\uff0c\u5219\u5176\u626b\u63cf\u8fc7\u7a0b\u7684\u6811\u5f62\u7ed3\u6784\u5982\u4e0b\uff1a<\/p>\n
\u56fe\u7247<\/p>\n
\u5728\u5176\u4e2d\u67d0\u4e2a\u8fc7\u7a0b\u4e2d\uff0c\u53ef\u80fd\u51fa\u73b0\u975e\u9891\u7e41\u7684\u9879\u76ee\u96c6\uff0c\u5c06\u5176\u53bb\u6389\uff08\u7528\u9634\u5f71\u8868\u793a\uff09\u4e3a\uff1a<\/p>\n
\u56fe\u7247<\/p>\n
<\/p>\n
FP Growth\u662f\u4e00\u79cd\u6bd4Apriori\u66f4\u9ad8\u6548\u7684\u9891\u7e41\u9879\u6316\u6398\u65b9\u6cd5\uff0c\u5b83\u53ea\u9700\u8981\u626b\u63cf\u9879\u76ee\u88682\u6b21\u3002\u5176\u4e2d\u7b2c1\u6b21\u626b\u63cf\u83b7\u5f97\u5f53\u4e2a\u9879\u76ee\u7684\u9891\u7387\uff0c\u53bb\u6389\u4e0d\u7b26\u5408\u652f\u6301\u5ea6\u8981\u6c42\u7684\u9879\uff0c\u5e76\u5bf9\u5269\u4e0b\u7684\u9879\u6392\u5e8f\u3002\u7b2c2\u904d\u626b\u63cf\u662f\u5efa\u7acb\u4e00\u9897FP-Tree(frequent-patten tree)\u3002<\/p>\n
<\/p>\n
\u63a5\u4e0b\u6765\u7684\u5de5\u4f5c\u5c31\u662f\u5728FP-Tree\u4e0a\u8fdb\u884c\u6316\u6398\uff0c\u6bd4\u5982\u8bf4\u6709\u4e0b\u8868\uff1a<\/p>\n
\u56fe\u7247<\/p>\n
\u5b83\u6240\u5bf9\u5e94\u7684FP_Tree\u5982\u4e0b\uff1a<\/p>\n
\u56fe\u7247<\/p>\n
<\/p>\n
\u7136\u540e\u4ece\u9891\u7387\u6700\u5c0f\u7684\u5355\u9879P\u5f00\u59cb\uff0c\u627e\u51faP\u7684\u6761\u4ef6\u6a21\u5f0f\u57fa\uff0c\u7528\u6784\u9020FP_Tree\u540c\u6837\u7684\u65b9\u6cd5\u6765\u6784\u9020P\u7684\u6761\u4ef6\u6a21\u5f0f\u57fa\u7684FP_Tree\uff0c\u5728\u8fd9\u68f5\u6811\u4e0a\u627e\u51fa\u5305\u542bP\u7684\u9891\u7e41\u9879\u96c6\u3002<\/p>\n
<\/p>\n
\u4f9d\u6b21\u4ecem,b,a,c,f\u7684\u6761\u4ef6\u6a21\u5f0f\u57fa\u4e0a\u6316\u6398\u9891\u7e41\u9879\u96c6\uff0c\u6709\u4e9b\u9879\u9700\u8981\u9012\u5f52\u7684\u53bb\u6316\u6398\uff0c\u6bd4\u8f83\u9ebb\u70e6\uff0c\u6bd4\u5982m\u8282\u70b9\u3002<\/p>\n