{"id":475945,"date":"2023-08-09T07:24:43","date_gmt":"2023-08-09T07:24:43","guid":{"rendered":""},"modified":"2023-09-05T11:11:40","modified_gmt":"2023-09-05T11:11:40","slug":"autoencoders","status":"publish","type":"wiki","link":"https:\/\/oneproxy.pro\/cn\/wiki\/autoencoders\/","title":{"rendered":"\u81ea\u52a8\u7f16\u7801\u5668"},"content":{"rendered":"<p>\u81ea\u52a8\u7f16\u7801\u5668\u662f\u4e00\u7c7b\u91cd\u8981\u4e14\u7528\u9014\u5e7f\u6cdb\u7684\u4eba\u5de5\u795e\u7ecf\u7f51\u7edc\uff0c\u4e3b\u8981\u7528\u4e8e\u65e0\u76d1\u7763\u5b66\u4e60\u4efb\u52a1\u3002\u5b83\u4eec\u4ee5\u5176\u6267\u884c\u964d\u7ef4\u3001\u7279\u5f81\u5b66\u4e60\u751a\u81f3\u751f\u6210\u5efa\u6a21\u7b49\u4efb\u52a1\u7684\u80fd\u529b\u800c\u95fb\u540d\u3002<\/p>\n<h2>\u81ea\u52a8\u7f16\u7801\u5668\u7684\u5386\u53f2<\/h2>\n<p>\u81ea\u52a8\u7f16\u7801\u5668\u7684\u6982\u5ff5\u8d77\u6e90\u4e8e 20 \u4e16\u7eaa 80 \u5e74\u4ee3\u968f\u7740 Hopfield \u7f51\u7edc\u7684\u53d1\u5c55\uff0c\u5b83\u662f\u73b0\u4ee3\u81ea\u52a8\u7f16\u7801\u5668\u7684\u5148\u9a71\u3002 Rumelhart \u7b49\u4eba\u4e8e 1986 \u5e74\u9996\u6b21\u63d0\u51fa\u81ea\u52a8\u7f16\u7801\u5668\u7684\u60f3\u6cd5\uff0c\u5f53\u65f6\u6b63\u503c\u4eba\u5de5\u795e\u7ecf\u7f51\u7edc\u7684\u65e9\u671f\u9636\u6bb5\u3002\u968f\u7740\u79d1\u5b66\u5bb6\u5f00\u59cb\u8ba4\u8bc6\u5230\u5b83\u4eec\u72ec\u7279\u7684\u81ea\u7f16\u7801\u80fd\u529b\uff0c\u201c\u81ea\u52a8\u7f16\u7801\u5668\u201d\u4e00\u8bcd\u540e\u6765\u88ab\u5efa\u7acb\u3002\u8fd1\u5e74\u6765\uff0c\u968f\u7740\u6df1\u5ea6\u5b66\u4e60\u7684\u5174\u8d77\uff0c\u81ea\u52a8\u7f16\u7801\u5668\u7ecf\u5386\u4e86\u590d\u5174\uff0c\u4e3a\u5f02\u5e38\u68c0\u6d4b\u3001\u964d\u566a\u751a\u81f3\u53d8\u5206\u81ea\u52a8\u7f16\u7801\u5668\uff08VAE\uff09\u7b49\u751f\u6210\u6a21\u578b\u505a\u51fa\u4e86\u91cd\u5927\u8d21\u732e\u3002<\/p>\n<h2>\u63a2\u7d22\u81ea\u52a8\u7f16\u7801\u5668<\/h2>\n<p>\u81ea\u52a8\u7f16\u7801\u5668\u662f\u4e00\u79cd\u4eba\u5de5\u795e\u7ecf\u7f51\u7edc\uff0c\u7528\u4e8e\u5b66\u4e60\u8f93\u5165\u6570\u636e\u7684\u6709\u6548\u7f16\u7801\u3002\u4e2d\u5fc3\u601d\u60f3\u662f\u5c06\u8f93\u5165\u7f16\u7801\u4e3a\u538b\u7f29\u8868\u793a\uff0c\u7136\u540e\u6839\u636e\u8be5\u8868\u793a\u5c3d\u53ef\u80fd\u51c6\u786e\u5730\u91cd\u5efa\u539f\u59cb\u8f93\u5165\u3002\u6b64\u8fc7\u7a0b\u6d89\u53ca\u4e24\u4e2a\u4e3b\u8981\u7ec4\u4ef6\uff1a\u7f16\u7801\u5668\uff08\u5c06\u8f93\u5165\u6570\u636e\u8f6c\u6362\u4e3a\u7d27\u51d1\u4ee3\u7801\uff09\u548c\u89e3\u7801\u5668\uff08\u4ece\u4ee3\u7801\u91cd\u5efa\u539f\u59cb\u8f93\u5165\uff09\u3002<\/p>\n<p>\u81ea\u52a8\u7f16\u7801\u5668\u7684\u76ee\u6807\u662f\u6700\u5c0f\u5316\u539f\u59cb\u8f93\u5165\u548c\u91cd\u6784\u8f93\u51fa\u4e4b\u95f4\u7684\u5dee\u5f02\uff08\u6216\u8bef\u5dee\uff09\uff0c\u4ece\u800c\u5b66\u4e60\u6570\u636e\u4e2d\u6700\u672c\u8d28\u7684\u7279\u5f81\u3002\u81ea\u52a8\u7f16\u7801\u5668\u5b66\u4e60\u7684\u538b\u7f29\u4ee3\u7801\u901a\u5e38\u5177\u6709\u6bd4\u539f\u59cb\u6570\u636e\u4f4e\u5f97\u591a\u7684\u7ef4\u5ea6\uff0c\u5bfc\u81f4\u81ea\u52a8\u7f16\u7801\u5668\u5728\u964d\u7ef4\u4efb\u52a1\u4e2d\u5e7f\u6cdb\u4f7f\u7528\u3002<\/p>\n<h2>\u81ea\u52a8\u7f16\u7801\u5668\u7684\u5185\u90e8\u7ed3\u6784<\/h2>\n<p>\u81ea\u52a8\u7f16\u7801\u5668\u7684\u67b6\u6784\u7531\u4e09\u4e2a\u4e3b\u8981\u90e8\u5206\u7ec4\u6210\uff1a<\/p>\n<ol>\n<li>\n<p><strong>\u7f16\u7801\u5668\uff1a<\/strong> \u7f51\u7edc\u7684\u8fd9\u4e00\u90e8\u5206\u5c06\u8f93\u5165\u538b\u7f29\u4e3a\u6f5c\u5728\u7a7a\u95f4\u8868\u793a\u3002\u5b83\u5c06\u8f93\u5165\u56fe\u50cf\u7f16\u7801\u4e3a\u964d\u7ef4\u7684\u538b\u7f29\u8868\u793a\u3002\u538b\u7f29\u56fe\u50cf\u901a\u5e38\u4fdd\u5b58\u6709\u5173\u8f93\u5165\u56fe\u50cf\u7684\u5173\u952e\u4fe1\u606f\u3002<\/p>\n<\/li>\n<li>\n<p><strong>\u74f6\u9888\uff1a<\/strong> \u8be5\u5c42\u4f4d\u4e8e\u7f16\u7801\u5668\u548c\u89e3\u7801\u5668\u4e4b\u95f4\u3002\u5b83\u5305\u542b\u8f93\u5165\u6570\u636e\u7684\u538b\u7f29\u8868\u793a\u3002\u8fd9\u662f\u8f93\u5165\u6570\u636e\u7684\u6700\u4f4e\u53ef\u80fd\u7ef4\u5ea6\u3002<\/p>\n<\/li>\n<li>\n<p><strong>\u89e3\u7801\u5668\uff1a<\/strong> \u7f51\u7edc\u7684\u8fd9\u4e00\u90e8\u5206\u6839\u636e\u7f16\u7801\u5f62\u5f0f\u91cd\u5efa\u8f93\u5165\u56fe\u50cf\u3002\u91cd\u5efa\u5c06\u662f\u539f\u59cb\u8f93\u5165\u7684\u6709\u635f\u91cd\u5efa\uff0c\u7279\u522b\u662f\u5f53\u7f16\u7801\u7ef4\u5ea6\u5c0f\u4e8e\u8f93\u5165\u7ef4\u5ea6\u65f6\u3002<\/p>\n<\/li>\n<\/ol>\n<p>\u6bcf\u4e2a\u90e8\u5206\u90fd\u7531\u591a\u5c42\u795e\u7ecf\u5143\u7ec4\u6210\uff0c\u5177\u4f53\u67b6\u6784\uff08\u5c42\u6570\u3001\u6bcf\u5c42\u795e\u7ecf\u5143\u6570\u91cf\u7b49\uff09\u53ef\u80fd\u6839\u636e\u5e94\u7528\u800c\u6709\u5f88\u5927\u5dee\u5f02\u3002<\/p>\n<h2>\u81ea\u52a8\u7f16\u7801\u5668\u7684\u4e3b\u8981\u7279\u70b9<\/h2>\n<ul>\n<li>\n<p><strong>\u6570\u636e\u7279\u5b9a\uff1a<\/strong> \u81ea\u52a8\u7f16\u7801\u5668\u88ab\u8bbe\u8ba1\u4e3a\u7279\u5b9a\u4e8e\u6570\u636e\u7684\uff0c\u8fd9\u610f\u5473\u7740\u5b83\u4eec\u4e0d\u4f1a\u5bf9\u672a\u7ecf\u8bad\u7ec3\u7684\u6570\u636e\u8fdb\u884c\u7f16\u7801\u3002<\/p>\n<\/li>\n<li>\n<p><strong>\u6709\u635f\uff1a<\/strong> \u8f93\u5165\u6570\u636e\u7684\u91cd\u5efa\u5c06\u662f\u201c\u6709\u635f\u7684\u201d\uff0c\u8fd9\u610f\u5473\u7740\u5728\u7f16\u7801\u8fc7\u7a0b\u4e2d\u603b\u4f1a\u4e22\u5931\u4e00\u4e9b\u4fe1\u606f\u3002<\/p>\n<\/li>\n<li>\n<p><strong>\u65e0\u76d1\u7763\uff1a<\/strong> \u81ea\u52a8\u7f16\u7801\u5668\u662f\u4e00\u79cd\u65e0\u76d1\u7763\u5b66\u4e60\u6280\u672f\uff0c\u56e0\u4e3a\u5b83\u4eec\u4e0d\u9700\u8981\u660e\u786e\u7684\u6807\u7b7e\u6765\u5b66\u4e60\u8868\u793a\u3002<\/p>\n<\/li>\n<li>\n<p><strong>\u964d\u7ef4\uff1a<\/strong> \u5b83\u4eec\u901a\u5e38\u7528\u4e8e\u964d\u7ef4\uff0c\u901a\u8fc7\u5b66\u4e60\u975e\u7ebf\u6027\u53d8\u6362\uff0c\u5176\u6027\u80fd\u4f18\u4e8e PCA \u7b49\u6280\u672f\u3002<\/p>\n<\/li>\n<\/ul>\n<h2>\u81ea\u52a8\u7f16\u7801\u5668\u7684\u7c7b\u578b<\/h2>\n<p>\u81ea\u52a8\u7f16\u7801\u5668\u6709\u591a\u79cd\u7c7b\u578b\uff0c\u6bcf\u79cd\u90fd\u6709\u5176\u72ec\u7279\u7684\u7279\u6027\u548c\u7528\u9014\u3002\u4ee5\u4e0b\u662f\u4e00\u4e9b\u5e38\u89c1\u7684\uff1a<\/p>\n<ol>\n<li>\n<p><strong>\u666e\u901a\u81ea\u52a8\u7f16\u7801\u5668\uff1a<\/strong> \u81ea\u52a8\u7f16\u7801\u5668\u6700\u7b80\u5355\u7684\u5f62\u5f0f\u662f\u524d\u9988\u3001\u975e\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\uff0c\u7c7b\u4f3c\u4e8e\u591a\u5c42\u611f\u77e5\u5668\u3002<\/p>\n<\/li>\n<li>\n<p><strong>\u591a\u5c42\u81ea\u52a8\u7f16\u7801\u5668\uff1a<\/strong> \u5982\u679c\u81ea\u52a8\u7f16\u7801\u5668\u4f7f\u7528\u591a\u4e2a\u9690\u85cf\u5c42\u8fdb\u884c\u7f16\u7801\u548c\u89e3\u7801\u8fc7\u7a0b\uff0c\u5219\u5b83\u88ab\u89c6\u4e3a\u591a\u5c42\u81ea\u52a8\u7f16\u7801\u5668\u3002<\/p>\n<\/li>\n<li>\n<p><strong>\u5377\u79ef\u81ea\u52a8\u7f16\u7801\u5668\uff1a<\/strong> \u8fd9\u4e9b\u81ea\u52a8\u7f16\u7801\u5668\u4f7f\u7528\u5377\u79ef\u5c42\u800c\u4e0d\u662f\u5168\u8fde\u63a5\u5c42\uff0c\u5e76\u4e0e\u56fe\u50cf\u6570\u636e\u4e00\u8d77\u4f7f\u7528\u3002<\/p>\n<\/li>\n<li>\n<p><strong>\u7a00\u758f\u81ea\u52a8\u7f16\u7801\u5668\uff1a<\/strong> \u8fd9\u4e9b\u81ea\u52a8\u7f16\u7801\u5668\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u5bf9\u9690\u85cf\u5355\u5143\u65bd\u52a0\u7a00\u758f\u6027\uff0c\u4ee5\u5b66\u4e60\u66f4\u7a33\u5065\u7684\u7279\u5f81\u3002<\/p>\n<\/li>\n<li>\n<p><strong>\u53bb\u566a\u81ea\u52a8\u7f16\u7801\u5668\uff1a<\/strong> \u8fd9\u4e9b\u81ea\u52a8\u7f16\u7801\u5668\u7ecf\u8fc7\u8bad\u7ec3\uff0c\u53ef\u4ee5\u4ece\u635f\u574f\u7684\u7248\u672c\u4e2d\u91cd\u5efa\u8f93\u5165\uff0c\u4ece\u800c\u6709\u52a9\u4e8e\u964d\u4f4e\u566a\u58f0\u3002<\/p>\n<\/li>\n<li>\n<p><strong>\u53d8\u5206\u81ea\u52a8\u7f16\u7801\u5668\uff08VAE\uff09\uff1a<\/strong> VAE \u662f\u4e00\u79cd\u81ea\u52a8\u7f16\u7801\u5668\uff0c\u53ef\u751f\u6210\u8fde\u7eed\u7684\u3001\u7ed3\u6784\u5316\u7684\u6f5c\u5728\u7a7a\u95f4\uff0c\u8fd9\u5bf9\u4e8e\u751f\u6210\u5efa\u6a21\u975e\u5e38\u6709\u7528\u3002<\/p>\n<\/li>\n<\/ol>\n<table>\n<thead>\n<tr>\n<th>\u81ea\u52a8\u7f16\u7801\u5668\u7c7b\u578b<\/th>\n<th>\u7279\u5f81<\/th>\n<th>\u5178\u578b\u7528\u4f8b<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>\u9999\u8349<\/td>\n<td>\u6700\u7b80\u5355\u7684\u5f62\u5f0f\uff0c\u7c7b\u4f3c\u4e8e\u591a\u5c42\u611f\u77e5\u5668<\/td>\n<td>\u57fa\u672c\u964d\u7ef4<\/td>\n<\/tr>\n<tr>\n<td>\u591a\u5c42<\/td>\n<td>\u7528\u4e8e\u7f16\u7801\u548c\u89e3\u7801\u7684\u591a\u4e2a\u9690\u85cf\u5c42<\/td>\n<td>\u590d\u6570\u964d\u7ef4<\/td>\n<\/tr>\n<tr>\n<td>\u5377\u79ef<\/td>\n<td>\u4f7f\u7528\u5377\u79ef\u5c42\uff0c\u901a\u5e38\u4e0e\u56fe\u50cf\u6570\u636e\u4e00\u8d77\u4f7f\u7528<\/td>\n<td>\u56fe\u50cf\u8bc6\u522b\u3001\u56fe\u50cf\u964d\u566a<\/td>\n<\/tr>\n<tr>\n<td>\u758f<\/td>\n<td>\u5bf9\u9690\u85cf\u5355\u5143\u65bd\u52a0\u7a00\u758f\u6027<\/td>\n<td>\u7279\u5f81\u9009\u62e9<\/td>\n<\/tr>\n<tr>\n<td>\u53bb\u566a<\/td>\n<td>\u7ecf\u8fc7\u8bad\u7ec3\uff0c\u53ef\u4ee5\u4ece\u635f\u574f\u7684\u7248\u672c\u4e2d\u91cd\u5efa\u8f93\u5165<\/td>\n<td>\u964d\u566a<\/td>\n<\/tr>\n<tr>\n<td>\u53d8\u5206\u6cd5<\/td>\n<td>\u4ea7\u751f\u8fde\u7eed\u7684\u3001\u7ed3\u6784\u5316\u7684\u6f5c\u5728\u7a7a\u95f4<\/td>\n<td>\u751f\u6210\u5efa\u6a21<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<h2>\u4f7f\u7528\u81ea\u52a8\u7f16\u7801\u5668\uff1a\u5e94\u7528\u548c\u6311\u6218<\/h2>\n<p>\u81ea\u52a8\u7f16\u7801\u5668\u5728\u673a\u5668\u5b66\u4e60\u548c\u6570\u636e\u5206\u6790\u4e2d\u6709\u8bb8\u591a\u5e94\u7528\uff1a<\/p>\n<ol>\n<li>\n<p><strong>\u6570\u636e\u538b\u7f29\uff1a<\/strong> \u53ef\u4ee5\u8bad\u7ec3\u81ea\u52a8\u7f16\u7801\u5668\u4ee5\u53ef\u4ee5\u5b8c\u7f8e\u91cd\u5efa\u7684\u65b9\u5f0f\u538b\u7f29\u6570\u636e\u3002<\/p>\n<\/li>\n<li>\n<p><strong>\u56fe\u50cf\u7740\u8272\uff1a<\/strong> \u81ea\u52a8\u7f16\u7801\u5668\u53ef\u7528\u4e8e\u5c06\u9ed1\u767d\u56fe\u50cf\u8f6c\u6362\u4e3a\u5f69\u8272\u56fe\u50cf\u3002<\/p>\n<\/li>\n<li>\n<p><strong>\u5f02\u5e38\u68c0\u6d4b\uff1a<\/strong> \u901a\u8fc7\u5bf9\u201c\u6b63\u5e38\u201d\u6570\u636e\u8fdb\u884c\u8bad\u7ec3\uff0c\u81ea\u52a8\u7f16\u7801\u5668\u53ef\u7528\u4e8e\u901a\u8fc7\u6bd4\u8f83\u91cd\u5efa\u8bef\u5dee\u6765\u68c0\u6d4b\u5f02\u5e38\u3002<\/p>\n<\/li>\n<li>\n<p><strong>\u56fe\u50cf\u53bb\u566a\uff1a<\/strong> \u81ea\u52a8\u7f16\u7801\u5668\u53ef\u7528\u4e8e\u6d88\u9664\u56fe\u50cf\u4e2d\u7684\u566a\u58f0\uff0c\u8fd9\u4e00\u8fc7\u7a0b\u79f0\u4e3a\u53bb\u566a\u3002<\/p>\n<\/li>\n<li>\n<p><strong>\u751f\u6210\u65b0\u6570\u636e\uff1a<\/strong> \u53d8\u5206\u81ea\u52a8\u7f16\u7801\u5668\u53ef\u4ee5\u751f\u6210\u4e0e\u8bad\u7ec3\u6570\u636e\u5177\u6709\u76f8\u540c\u7edf\u8ba1\u6570\u636e\u7684\u65b0\u6570\u636e\u3002<\/p>\n<\/li>\n<\/ol>\n<p>\u7136\u800c\uff0c\u81ea\u52a8\u7f16\u7801\u5668\u4e5f\u53ef\u80fd\u5e26\u6765\u6311\u6218\uff1a<\/p>\n<ul>\n<li>\n<p>\u81ea\u52a8\u7f16\u7801\u5668\u53ef\u80fd\u5bf9\u8f93\u5165\u6570\u636e\u89c4\u6a21\u654f\u611f\u3002\u901a\u5e38\u9700\u8981\u8fdb\u884c\u7279\u5f81\u7f29\u653e\u624d\u80fd\u83b7\u5f97\u826f\u597d\u7684\u7ed3\u679c\u3002<\/p>\n<\/li>\n<li>\n<p>\u7406\u60f3\u7684\u67b6\u6784\uff08\u5373\u5c42\u6570\u548c\u6bcf\u5c42\u8282\u70b9\u6570\uff09\u662f\u9ad8\u5ea6\u9488\u5bf9\u7279\u5b9a\u95ee\u9898\u7684\uff0c\u5e76\u4e14\u901a\u5e38\u9700\u8981\u8fdb\u884c\u5927\u91cf\u5b9e\u9a8c\u3002<\/p>\n<\/li>\n<li>\n<p>\u4e0e PCA \u7b49\u6280\u672f\u4e0d\u540c\uff0c\u6240\u5f97\u7684\u538b\u7f29\u8868\u793a\u901a\u5e38\u4e0d\u5bb9\u6613\u89e3\u91ca\u3002<\/p>\n<\/li>\n<li>\n<p>\u81ea\u52a8\u7f16\u7801\u5668\u53ef\u80fd\u5bf9\u8fc7\u5ea6\u62df\u5408\u5f88\u654f\u611f\uff0c\u5c24\u5176\u662f\u5f53\u7f51\u7edc\u67b6\u6784\u5177\u6709\u9ad8\u5bb9\u91cf\u65f6\u3002<\/p>\n<\/li>\n<\/ul>\n<h2>\u6bd4\u8f83\u548c\u76f8\u5173\u6280\u672f<\/h2>\n<p>\u81ea\u52a8\u7f16\u7801\u5668\u53ef\u4ee5\u4e0e\u5176\u4ed6\u964d\u7ef4\u548c\u65e0\u76d1\u7763\u5b66\u4e60\u6280\u672f\u8fdb\u884c\u6bd4\u8f83\uff0c\u5982\u4e0b\u6240\u793a\uff1a<\/p>\n<table>\n<thead>\n<tr>\n<th>\u6280\u672f<\/th>\n<th>\u65e0\u76d1\u7763<\/th>\n<th>\u975e\u7ebf\u6027<\/th>\n<th>\u5185\u7f6e\u7279\u5f81\u9009\u62e9<\/th>\n<th>\u751f\u6210\u80fd\u529b<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>\u81ea\u52a8\u7f16\u7801\u5668<\/td>\n<td>\u662f\u7684<\/td>\n<td>\u662f\u7684<\/td>\n<td>\u662f\uff08\u7a00\u758f\u81ea\u52a8\u7f16\u7801\u5668\uff09<\/td>\n<td>\u662f\uff08VAE\uff09<\/td>\n<\/tr>\n<tr>\n<td>\u4e3b\u6210\u5206\u5206\u6790<\/td>\n<td>\u662f\u7684<\/td>\n<td>\u4e0d<\/td>\n<td>\u4e0d<\/td>\n<td>\u4e0d<\/td>\n<\/tr>\n<tr>\n<td>t-SNE<\/td>\n<td>\u662f\u7684<\/td>\n<td>\u662f\u7684<\/td>\n<td>\u4e0d<\/td>\n<td>\u4e0d<\/td>\n<\/tr>\n<tr>\n<td>K-\u5747\u503c\u805a\u7c7b<\/td>\n<td>\u662f\u7684<\/td>\n<td>\u4e0d<\/td>\n<td>\u4e0d<\/td>\n<td>\u4e0d<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<h2>\u81ea\u52a8\u7f16\u7801\u5668\u7684\u672a\u6765\u5c55\u671b<\/h2>\n<p>\u81ea\u52a8\u7f16\u7801\u5668\u6b63\u5728\u4e0d\u65ad\u5b8c\u5584\u548c\u6539\u8fdb\u3002\u672a\u6765\uff0c\u81ea\u52a8\u7f16\u7801\u5668\u9884\u8ba1\u5c06\u5728\u65e0\u76d1\u7763\u548c\u534a\u76d1\u7763\u5b66\u4e60\u3001\u5f02\u5e38\u68c0\u6d4b\u548c\u751f\u6210\u5efa\u6a21\u4e2d\u53d1\u6325\u66f4\u5927\u7684\u4f5c\u7528\u3002<\/p>\n<p>\u4e00\u4e2a\u4ee4\u4eba\u5174\u594b\u7684\u524d\u6cbf\u9886\u57df\u662f\u81ea\u52a8\u7f16\u7801\u5668\u4e0e\u5f3a\u5316\u5b66\u4e60 (RL) \u7684\u7ed3\u5408\u3002\u81ea\u52a8\u7f16\u7801\u5668\u53ef\u4ee5\u5e2e\u52a9\u5b66\u4e60\u73af\u5883\u7684\u6709\u6548\u8868\u793a\uff0c\u4f7f\u5f3a\u5316\u5b66\u4e60\u7b97\u6cd5\u66f4\u52a0\u9ad8\u6548\u3002\u6b64\u5916\uff0c\u81ea\u52a8\u7f16\u7801\u5668\u4e0e\u5176\u4ed6\u751f\u6210\u6a21\u578b\uff08\u4f8b\u5982\u751f\u6210\u5bf9\u6297\u7f51\u7edc\uff08GAN\uff09\uff09\u7684\u96c6\u6210\u662f\u521b\u5efa\u66f4\u5f3a\u5927\u7684\u751f\u6210\u6a21\u578b\u7684\u53e6\u4e00\u4e2a\u6709\u5e0c\u671b\u7684\u9014\u5f84\u3002<\/p>\n<h2>\u81ea\u52a8\u7f16\u7801\u5668\u548c\u4ee3\u7406\u670d\u52a1\u5668<\/h2>\n<p>\u81ea\u52a8\u7f16\u7801\u5668\u548c\u4ee3\u7406\u670d\u52a1\u5668\u4e4b\u95f4\u7684\u5173\u7cfb\u4e0d\u662f\u76f4\u63a5\u7684\uff0c\u800c\u662f\u4e3b\u8981\u662f\u4e0a\u4e0b\u6587\u76f8\u5173\u7684\u3002\u4ee3\u7406\u670d\u52a1\u5668\u4e3b\u8981\u5145\u5f53\u5ba2\u6237\u7aef\u4ece\u5176\u4ed6\u670d\u52a1\u5668\u5bfb\u6c42\u8d44\u6e90\u7684\u8bf7\u6c42\u7684\u4e2d\u4ecb\uff0c\u63d0\u4f9b\u9690\u79c1\u4fdd\u62a4\u3001\u8bbf\u95ee\u63a7\u5236\u548c\u7f13\u5b58\u7b49\u5404\u79cd\u529f\u80fd\u3002<\/p>\n<p>\u867d\u7136\u81ea\u52a8\u7f16\u7801\u5668\u7684\u4f7f\u7528\u53ef\u80fd\u4e0d\u4f1a\u76f4\u63a5\u589e\u5f3a\u4ee3\u7406\u670d\u52a1\u5668\u7684\u529f\u80fd\uff0c\u4f46\u53ef\u4ee5\u5728\u4ee3\u7406\u670d\u52a1\u5668\u662f\u7f51\u7edc\u4e00\u90e8\u5206\u7684\u5927\u578b\u7cfb\u7edf\u4e2d\u5229\u7528\u5b83\u4eec\u3002\u4f8b\u5982\uff0c\u5982\u679c\u4ee3\u7406\u670d\u52a1\u5668\u662f\u5904\u7406\u5927\u91cf\u6570\u636e\u7684\u7cfb\u7edf\u7684\u4e00\u90e8\u5206\uff0c\u5219\u81ea\u52a8\u7f16\u7801\u5668\u53ef\u7528\u4e8e\u6570\u636e\u538b\u7f29\u6216\u68c0\u6d4b\u7f51\u7edc\u6d41\u91cf\u4e2d\u7684\u5f02\u5e38\u3002<\/p>\n<p>\u53e6\u4e00\u4e2a\u6f5c\u5728\u7684\u5e94\u7528\u662f\u5728 VPN \u6216\u5176\u4ed6\u5b89\u5168\u4ee3\u7406\u670d\u52a1\u5668\u7684\u73af\u5883\u4e2d\uff0c\u5176\u4e2d\u81ea\u52a8\u7f16\u7801\u5668\u53ef\u80fd\u88ab\u7528\u4f5c\u68c0\u6d4b\u7f51\u7edc\u6d41\u91cf\u4e2d\u5f02\u5e38\u6216\u5f02\u5e38\u6a21\u5f0f\u7684\u673a\u5236\uff0c\u4ece\u800c\u6709\u52a9\u4e8e\u7f51\u7edc\u7684\u5b89\u5168\u3002<\/p>\n<h2>\u76f8\u5173\u94fe\u63a5<\/h2>\n<p>\u8981\u8fdb\u4e00\u6b65\u63a2\u7d22\u81ea\u52a8\u7f16\u7801\u5668\uff0c\u8bf7\u53c2\u9605\u4ee5\u4e0b\u8d44\u6e90\uff1a<\/p>\n<ol>\n<li>\n<p><a href=\"https:\/\/www.deeplearningbook.org\/contents\/autoencoders.html\" target=\"_new\" rel=\"noopener nofollow\">\u6df1\u5ea6\u5b66\u4e60\u4e2d\u7684\u81ea\u52a8\u7f16\u7801\u5668<\/a> \u2013 Goodfellow\u3001Bengio \u548c Courville \u7f16\u5199\u7684\u6df1\u5ea6\u5b66\u4e60\u6559\u79d1\u4e66\u3002<\/p>\n<\/li>\n<li>\n<p><a href=\"https:\/\/blog.keras.io\/building-autoencoders-in-keras.html\" target=\"_new\" rel=\"noopener nofollow\">\u5728 Keras \u4e2d\u6784\u5efa\u81ea\u52a8\u7f16\u7801\u5668<\/a> \u2013 \u5728 Keras \u4e2d\u5b9e\u73b0\u81ea\u52a8\u7f16\u7801\u5668\u7684\u6559\u7a0b\u3002<\/p>\n<\/li>\n<li>\n<p><a href=\"https:\/\/wiseodd.github.io\/techblog\/2016\/12\/10\/variational-autoencoder\/\" target=\"_new\" rel=\"noopener nofollow\">\u53d8\u5206\u81ea\u52a8\u7f16\u7801\u5668\uff1a\u76f4\u89c9\u548c\u5b9e\u73b0<\/a> \u2013 \u53d8\u5206\u81ea\u52a8\u7f16\u7801\u5668\u7684\u89e3\u91ca\u548c\u5b9e\u73b0\u3002<\/p>\n<\/li>\n<li>\n<p><a href=\"http:\/\/deeplearning.stanford.edu\/tutorial\/supervised\/FeatureExtractionUsingConvolution\/\" target=\"_new\" rel=\"noopener nofollow\">\u7a00\u758f\u81ea\u52a8\u7f16\u7801\u5668<\/a> \u2013 \u65af\u5766\u798f\u5927\u5b66\u5173\u4e8e\u7a00\u758f\u81ea\u52a8\u7f16\u7801\u5668\u7684\u6559\u7a0b\u3002<\/p>\n<\/li>\n<li>\n<p><a href=\"https:\/\/towardsdatascience.com\/understanding-variational-autoencoders-vaes-f70510919f73\" target=\"_new\" rel=\"noopener nofollow\">\u4e86\u89e3\u53d8\u5206\u81ea\u52a8\u7f16\u7801\u5668 (VAE)<\/a> \u2013 \u6765\u81ea\u8d70\u5411\u6570\u636e\u79d1\u5b66\u7684\u5173\u4e8e\u53d8\u5206\u81ea\u52a8\u7f16\u7801\u5668\u7684\u7efc\u5408\u6587\u7ae0\u3002<\/p>\n<\/li>\n<\/ol>","protected":false},"featured_media":467668,"menu_order":0,"template":"","meta":{"_acf_changed":false,"content-type":"","inline_featured_image":false,"footnotes":""},"class_list":["post-475945","wiki","type-wiki","status-publish","has-post-thumbnail","hentry"],"acf":{"faq_title":"Frequently Asked Questions about <mark>Autoencoders: Unsupervised Learning and Data Compression<\/mark>","faq_items":[{"question":"What are Autoencoders?","answer":"<p>Autoencoders are a class of artificial neural networks used primarily for unsupervised learning tasks. They function by encoding input data into a compressed representation and then reconstructing the original input as accurately as possible from this representation. This process involves two primary components: an encoder and a decoder. Autoencoders are particularly useful for tasks such as dimensionality reduction, feature learning, and generative modeling.<\/p>"},{"question":"What is the history of Autoencoders?","answer":"<p>The concept of autoencoders originated in the 1980s with the development of the Hopfield Network. The term 'autoencoder' came into use as scientists started recognizing the unique self-encoding capabilities of these networks. Over the years, particularly with the advent of deep learning, autoencoders have found extensive use in areas like anomaly detection, noise reduction, and generative models.<\/p>"},{"question":"How does an Autoencoder work?","answer":"<p>An autoencoder works by encoding the input data into a compressed representation and then reconstructing the original input from this representation. This process involves two main components: an encoder, which transforms the input data into a compact code, and a decoder, which reconstructs the original input from the code. The objective of an autoencoder is to minimize the difference (or error) between the original input and the reconstructed output.<\/p>"},{"question":"What are the key features of Autoencoders?","answer":"<p>Autoencoders are data-specific, implying that they won't encode data for which they were not trained. They are also lossy, meaning that some information is always lost in the encoding process. Autoencoders are an unsupervised learning technique as they do not require explicit labels to learn the representation. Finally, they are often used for dimensionality reduction, where they can learn non-linear transformations of the data.<\/p>"},{"question":"What are the different types of Autoencoders?","answer":"<p>Several types of autoencoders exist, including Vanilla Autoencoder, Multilayer Autoencoder, Convolutional Autoencoder, Sparse Autoencoder, Denoising Autoencoder, and Variational Autoencoder (VAE). Each type of autoencoder has its unique characteristics and applications, ranging from basic dimensionality reduction to complex tasks like image recognition, feature selection, noise reduction, and generative modeling.<\/p>"},{"question":"How are Autoencoders used?","answer":"<p>Autoencoders have several applications, including data compression, image colorization, anomaly detection, denoising images, and generating new data. However, they can also pose challenges such as sensitivity to input data scale, difficulty determining the ideal architecture, the lack of interpretability of the compressed representation, and susceptibility to overfitting.<\/p>"},{"question":"How do Autoencoders compare with other techniques?","answer":"<p>Autoencoders are compared with other dimensionality reduction and unsupervised learning techniques based on several factors, including whether the technique is unsupervised, its ability to learn non-linear transformations, in-built feature selection capabilities, and whether it has generative capabilities. Compared to techniques like PCA, t-SNE, and K-means clustering, autoencoders often offer superior flexibility and performance, particularly in tasks involving non-linear transformations and generative modeling.<\/p>"},{"question":"What are the future perspectives on Autoencoders?","answer":"<p>Autoencoders are expected to play a significant role in future unsupervised and semi-supervised learning, anomaly detection, and generative modeling. Combining autoencoders with reinforcement learning or other generative models like Generative Adversarial Networks (GANs) is a promising avenue for creating more powerful generative models.<\/p>"},{"question":"How can Autoencoders be used with Proxy Servers?","answer":"<p>While autoencoders do not directly enhance the capabilities of a proxy server, they can be useful in systems where a proxy server is part of the network. Autoencoders can be used for data compression or for detecting anomalies in network traffic in such systems. Additionally, in the context of VPNs or other secure proxy servers, autoencoders could potentially be used to detect unusual or anomalous patterns in network traffic.<\/p>"}]},"_links":{"self":[{"href":"https:\/\/oneproxy.pro\/cn\/wp-json\/wp\/v2\/wiki\/475945","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/oneproxy.pro\/cn\/wp-json\/wp\/v2\/wiki"}],"about":[{"href":"https:\/\/oneproxy.pro\/cn\/wp-json\/wp\/v2\/types\/wiki"}],"version-history":[{"count":0,"href":"https:\/\/oneproxy.pro\/cn\/wp-json\/wp\/v2\/wiki\/475945\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/oneproxy.pro\/cn\/wp-json\/wp\/v2\/media\/467668"}],"wp:attachment":[{"href":"https:\/\/oneproxy.pro\/cn\/wp-json\/wp\/v2\/media?parent=475945"}],"curies":[{"name":"\u53ef\u6e7f\u6027\u7c89\u5242","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}