{"id":808,"date":"2025-12-30T15:43:54","date_gmt":"2025-12-30T07:43:54","guid":{"rendered":"https:\/\/www.ndnlab.com\/?p=808"},"modified":"2025-12-30T15:43:55","modified_gmt":"2025-12-30T07:43:55","slug":"hattrick-solving-multi-class-te-using-neural-models","status":"publish","type":"post","link":"https:\/\/www.ndnlab.com\/?p=808","title":{"rendered":"Hattrick: Solving Multi-Class TE using Neural Models"},"content":{"rendered":"\n<p class=\"wp-block-paragraph\">SIGCOMM &#8217;25: ACM SIGCOMM 2025 Conference <em>September 8 &#8211; 11, 2025<\/em> <em>Coimbra, Portugal<\/em><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Conference Sponsors:<\/strong> SIGCOMM  <\/p>\n\n\n\n<p class=\"wp-block-paragraph\">https:\/\/dl.acm.org\/doi\/10.1145\/3718958.3750470<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">\u4e00\u3001\u7814\u7a76\u80cc\u666f\u4e0e\u52a8\u673a\uff1a\u591a\u4f18\u5148\u7ea7\u6d41\u91cf\u5de5\u7a0b\u7684\u5174\u8d77<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">\u5728\u73b0\u4ee3\u8d85\u5927\u89c4\u6a21\u5e7f\u57df\u7f51\uff08WAN\uff09\u4e2d\uff0c\u6d41\u91cf\u4e0d\u518d\u88ab\u89c6\u4e3a\u5355\u4e00\u7c7b\u522b\uff0c\u800c\u662f\u6839\u636e\u4e1a\u52a1\u9700\u6c42\u88ab\u5212\u5206\u4e3a\u4e0d\u540c\u7684\u4f18\u5148\u7ea7\uff08\u5982\u9ad8\u3001\u4e2d\u3001\u4f4e\uff09\u3002\u4f8b\u5982\uff0c\u89c6\u9891\u4f1a\u8bae\u548c\u5b9e\u65f6\u63a7\u5236\u4fe1\u4ee4\u5c5e\u4e8e\u9ad8\u4f18\u5148\u7ea7\uff0c\u800c\u540e\u53f0\u6570\u636e\u5907\u4efd\u5219\u5c5e\u4e8e\u4f4e\u4f18\u5148\u7ea7\u3002\u8fd9\u79cd\u591a\u4f18\u5148\u7ea7\u6d41\u91cf\u5de5\u7a0b\uff08Multi-Class TE, MC-TE\uff09\u7684\u6838\u5fc3\u76ee\u6807\u662f\u5728\u786e\u4fdd\u9ad8\u4f18\u5148\u7ea7\u6d41\u91cf\u7edd\u5bf9\u4f18\u5148\u7684\u524d\u63d0\u4e0b\uff0c\u6700\u5927\u5316\u7f51\u7edc\u8d44\u6e90\u7684\u6574\u4f53\u5229\u7528\u7387\u3002<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">\u7136\u800c\uff0c\u73b0\u6709\u7684\u6d41\u91cf\u5de5\u7a0b\u65b9\u6848\u9762\u4e34\u4e25\u5cfb\u6311\u6218\u3002\u4f20\u7edf\u7684\u7ebf\u6027\u89c4\u5212\uff08LP\uff09\u65b9\u6cd5\u867d\u7136\u80fd\u6c42\u5f97\u6700\u4f18\u89e3\uff0c\u4f46\u5728\u9762\u5bf9\u5927\u89c4\u6a21\u7f51\u7edc\u62d3\u6251\u65f6\uff0c\u8ba1\u7b97\u8017\u65f6\u901a\u5e38\u8fbe\u5230\u5206\u949f\u7ea7\uff0c\u65e0\u6cd5\u9002\u5e94\u7f51\u7edc\u72b6\u6001\u7684\u77ac\u65f6\u6ce2\u52a8\u3002\u800c\u8fd1\u5e74\u6765\u5174\u8d77\u7684\u57fa\u4e8e\u6df1\u5ea6\u5b66\u4e60\u7684\u6d41\u91cf\u5de5\u7a0b\u6a21\u578b\uff0c\u5927\u591a\u4ec5\u9488\u5bf9\u5355\u4e00\u7c7b\u522b\u6d41\u91cf\u8bbe\u8ba1\uff0c\u96be\u4ee5\u5904\u7406\u591a\u4f18\u5148\u7ea7\u4e4b\u95f4\u4e25\u683c\u7684\u5c42\u6b21\u7ea6\u675f\u548c\u590d\u6742\u7684\u4f18\u5148\u7ea7\u62a2\u5360\u903b\u8f91\u3002<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">\u4e8c\u3001\u6838\u5fc3\u6311\u6218\uff1a\u4e25\u82db\u7ea6\u675f\u4e0b\u7684\u795e\u7ecf\u6a21\u578b\u8bbe\u8ba1<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">\u5c06\u795e\u7ecf\u7f51\u7edc\u5e94\u7528\u4e8e MC-TE \u9762\u4e34\u4e24\u4e2a\u4e3b\u8981\u74f6\u9888\uff1a\u9996\u5148\u662f<strong>\u4e25\u683c\u7684\u4f18\u5148\u7ea7\u4fdd\u969c<\/strong>\u3002\u5728\u9ad8\u8d1f\u8f7d\u60c5\u51b5\u4e0b\uff0c\u7cfb\u7edf\u5fc5\u987b\u786e\u4fdd\u4f4e\u4f18\u5148\u7ea7\u6d41\u91cf\u4e0d\u4f1a\u6324\u5360\u9ad8\u4f18\u5148\u7ea7\u6d41\u91cf\u7684\u5e26\u5bbd\uff0c\u8fd9\u79cd\u201c\u786c\u7ea6\u675f\u201d\u5728\u7aef\u5230\u7aef\u7684\u795e\u7ecf\u6a21\u578b\u4e2d\u6781\u96be\u5b9e\u73b0\u3002\u5176\u6b21\u662f<strong>\u6cdb\u5316\u6027\u4e0e\u89c4\u6a21\u5316\u95ee\u9898<\/strong>\u3002\u751f\u4ea7\u73af\u5883\u4e2d\u7684\u62d3\u6251\u7ed3\u6784\u548c\u6d41\u91cf\u6a21\u5f0f\uff08TM\uff09\u5904\u4e8e\u52a8\u6001\u53d8\u5316\u4e2d\uff0c\u795e\u7ecf\u6a21\u578b\u5fc5\u987b\u5728\u4fdd\u8bc1\u8ba1\u7b97\u6548\u7387\u7684\u540c\u65f6\uff0c\u5177\u5907\u8de8\u62d3\u6251\u7684\u9c81\u68d2\u6027\u3002<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">\u73b0\u6709\u7684\u795e\u7ecf\u6a21\u578b\u5728\u5904\u7406 MC-TE \u65f6\uff0c\u5f80\u5f80\u901a\u8fc7\u7b80\u5355\u7684\u52a0\u6743\u5956\u52b1\u51fd\u6570\u6765\u533a\u5206\u4f18\u5148\u7ea7\uff0c\u8fd9\u5728\u5b9e\u9645\u90e8\u7f72\u4e2d\u4f1a\u5bfc\u81f4\u9ad8\u4f18\u5148\u7ea7\u6d41\u91cf\u7684\u6027\u80fd\u53d7\u635f\u3002\u6b64\u5916\uff0c\u591a\u4f18\u5148\u7ea7\u4efb\u52a1\u4f1a\u5bfc\u81f4\u89e3\u7a7a\u95f4\u5448\u6307\u6570\u7ea7\u589e\u957f\uff0c\u666e\u901a\u7684\u56fe\u795e\u7ecf\u7f51\u7edc\uff08GNN\uff09\u67b6\u6784\u96be\u4ee5\u5728\u6709\u9650\u7684\u63a8\u7406\u65f6\u95f4\u5185\u6536\u655b\u5230\u80fd\u4e0e LP \u6700\u4f18\u89e3\u5ab2\u7f8e\u7684\u5206\u914d\u65b9\u6848\u3002<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">\u4e09\u3001Hattrick \u7cfb\u7edf\u67b6\u6784\uff1a\u591a\u9636\u6bb5\u795e\u7ecf\u8c03\u5ea6<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Hattrick<\/strong> \u63d0\u51fa\u4e86\u4e00\u79cd\u521b\u65b0\u7684<strong>\u591a\u9636\u6bb5\u5e76\u884c\u795e\u7ecf\u67b6\u6784<\/strong>\uff0c\u5176\u8bbe\u8ba1\u7075\u611f\u6765\u6e90\u4e8e\u89e3\u51b3 MC-TE \u7684\u591a\u9636\u6bb5\u7ebf\u6027\u89c4\u5212\u601d\u8def\u3002\u7cfb\u7edf\u5c06\u590d\u6742\u7684\u8c03\u5ea6\u95ee\u9898\u5206\u89e3\u4e3a\u591a\u4e2a\u5b50\u4efb\u52a1\uff0c\u6bcf\u4e2a\u9636\u6bb5\u5bf9\u5e94\u4e00\u4e2a\u6d41\u91cf\u7b49\u7ea7\u3002\u7b2c\u4e00\u9636\u6bb5\u4e13\u95e8\u8d1f\u8d23\u4e3a\u9ad8\u4f18\u5148\u7ea7\u6d41\u91cf\u5206\u914d\u8def\u5f84\uff1b\u968f\u540e\uff0c\u5728\u56fa\u5b9a\u9ad8\u4f18\u5148\u7ea7\u5206\u914d\u7ed3\u679c\u7684\u57fa\u7840\u4e0a\uff0c\u540e\u7eed\u9636\u6bb5\u4f9d\u6b21\u4e3a\u4e2d\u3001\u4f4e\u4f18\u5148\u7ea7\u6d41\u91cf\u5bfb\u627e\u5269\u4f59\u5e26\u5bbd\u4e2d\u7684\u6700\u4f18\u8def\u5f84\u3002<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">\u8be5\u67b6\u6784\u7684\u6838\u5fc3\u662f\u57fa\u4e8e\u56fe\u795e\u7ecf\u7f51\u7edc\uff08GNN\uff09\u7684\u7f16\u7801-\u89e3\u7801\u7ed3\u6784\u3002\u5b83\u80fd\u591f\u6355\u6349\u7f51\u7edc\u62d3\u6251\u4e2d\u94fe\u8def\u4e0e\u8282\u70b9\u4e4b\u95f4\u7684\u590d\u6742\u7a7a\u95f4\u4f9d\u8d56\u5173\u7cfb\u3002Hattrick \u901a\u8fc7\u8fd9\u79cd\u591a\u9636\u6bb5\u7684\u8bbe\u8ba1\uff0c\u4e0d\u4ec5\u5728\u903b\u8f91\u4e0a\u6a21\u62df\u4e86\u4f18\u5148\u7ea7\u7684\u5c42\u6b21\u611f\uff0c\u8fd8\u663e\u8457\u964d\u4f4e\u4e86\u5355\u4e2a\u6a21\u578b\u7684\u5b66\u4e60\u96be\u5ea6\u3002\u901a\u8fc7\u8fd9\u79cd\u201c\u5206\u800c\u6cbb\u4e4b\u201d\u7684\u7b56\u7565\uff0cHattrick \u6210\u529f\u5730\u5c06\u591a\u4f18\u5148\u7ea7\u7ea6\u675f\u5185\u5316\u5230\u4e86\u6a21\u578b\u7684\u7ed3\u6784\u4e2d\uff0c\u800c\u975e\u4ec5\u4ec5\u4f9d\u9760\u635f\u5931\u51fd\u6570\u8fdb\u884c\u5f15\u5bfc\u3002<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"374\" src=\"https:\/\/www.ndnlab.com\/wp-content\/uploads\/2025\/12\/1-1-1024x374.png\"  class=\"wp-image-810\" srcset=\"https:\/\/www.ndnlab.com\/wp-content\/uploads\/2025\/12\/1-1-1024x374.png 1024w, https:\/\/www.ndnlab.com\/wp-content\/uploads\/2025\/12\/1-1-300x110.png 300w, https:\/\/www.ndnlab.com\/wp-content\/uploads\/2025\/12\/1-1-768x280.png 768w, https:\/\/www.ndnlab.com\/wp-content\/uploads\/2025\/12\/1-1-1536x561.png 1536w, https:\/\/www.ndnlab.com\/wp-content\/uploads\/2025\/12\/1-1.png 1550w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" title=\"Hattrick: Solving Multi-Class TE using Neural Models\u63d2\u56fe\" alt=\"Hattrick: Solving Multi-Class TE using Neural Models\u63d2\u56fe\" \/><\/figure>\n\n\n\n<h2 class=\"wp-block-heading\">\u56db\u3001\u5173\u952e\u521b\u65b0\u70b9\uff1a\u5206\u5c42\u635f\u5931\u4e0e\u5feb\u901f\u63a8\u7406<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">Hattrick \u7684\u4e00\u5927\u6280\u672f\u7a81\u7834\u5728\u4e8e\u5176<strong>\u5206\u5c42\u635f\u5931\u51fd\u6570\uff08Hierarchical Loss Function\uff09<\/strong>\u3002\u4e3a\u4e86\u786e\u4fdd\u6a21\u578b\u80fd\u591f\u6536\u655b\u5230\u7b26\u5408\u4f18\u5148\u7ea7\u89c4\u5219\u7684\u89e3\uff0c\u56e2\u961f\u8bbe\u8ba1\u4e86\u4e00\u5957\u80fd\u591f\u53cd\u6620 MC-TE \u76ee\u6807\u7684\u590d\u5408\u5956\u52b1\u673a\u5236\u3002\u5b83\u5f3a\u5236\u8981\u6c42\u6a21\u578b\u5728\u4f18\u5316\u6574\u4f53\u541e\u5410\u91cf\u7684\u540c\u65f6\uff0c\u5bf9\u8fdd\u53cd\u9ad8\u4f18\u5148\u7ea7\u5e26\u5bbd\u4fdd\u969c\u7684\u884c\u4e3a\u65bd\u52a0\u4e25\u5389\u60e9\u7f5a\u3002\u8fd9\u79cd\u8bbe\u8ba1\u4f7f\u5f97\u6a21\u578b\u5728\u65e0\u76d1\u7763\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\uff0c\u80fd\u591f\u81ea\u52a8\u4e60\u5f97\u590d\u6742\u7684\u8def\u5f84\u5207\u6362\u7b56\u7565\uff0c\u4ee5\u5e94\u5bf9\u94fe\u8def\u6545\u969c\u6216\u6d41\u91cf\u6fc0\u589e\u3002<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">\u5728\u6027\u80fd\u63d0\u5347\u65b9\u9762\uff0cHattrick \u5b9e\u73b0\u4e86<strong>\u8fd1\u4e4e\u5b9e\u65f6\u7684\u63a8\u7406\u901f\u5ea6<\/strong>\u3002\u4e0e\u4f20\u7edf\u7684 LP \u6c42\u89e3\u5668\u76f8\u6bd4\uff0cHattrick \u7684\u51b3\u7b56\u901f\u5ea6\u63d0\u5347\u4e86 <strong>100 \u500d\u4ee5\u4e0a<\/strong>\uff0c\u5c06\u539f\u672c\u9700\u8981\u5206\u949f\u7ea7\u7684\u8ba1\u7b97\u7f29\u77ed\u81f3\u79d2\u7ea7\u751a\u81f3\u6beb\u79d2\u7ea7\u3002\u8fd9\u79cd\u6781\u81f4\u7684\u6548\u7387\u4f7f\u5f97\u7f51\u7edc\u7ba1\u7406\u5458\u80fd\u591f\u4ee5\u66f4\u9ad8\u7684\u9891\u7387\u8fdb\u884c\u6d41\u91cf\u91cd\u8def\u7531\uff08Re-routing\uff09\uff0c\u4ece\u800c\u6781\u5927\u5730\u589e\u5f3a\u4e86\u5e7f\u57df\u7f51\u5e94\u5bf9\u4e9a\u79d2\u7ea7\u7a81\u53d1\u6d41\u91cf\u7684\u97e7\u6027\u3002<\/p>\n\n\n\n<figure class=\"wp-block-image size-full\"><img loading=\"lazy\" decoding=\"async\" width=\"522\" height=\"258\" src=\"https:\/\/www.ndnlab.com\/wp-content\/uploads\/2025\/12\/\u516c\u5f0f.png\"  class=\"wp-image-815\" srcset=\"https:\/\/www.ndnlab.com\/wp-content\/uploads\/2025\/12\/\u516c\u5f0f.png 522w, https:\/\/www.ndnlab.com\/wp-content\/uploads\/2025\/12\/\u516c\u5f0f-300x148.png 300w\" sizes=\"auto, (max-width: 522px) 100vw, 522px\" title=\"Hattrick: Solving Multi-Class TE using Neural Models\u63d2\u56fe1\" alt=\"Hattrick: Solving Multi-Class TE using Neural Models\u63d2\u56fe1\" \/><\/figure>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"365\" src=\"https:\/\/www.ndnlab.com\/wp-content\/uploads\/2025\/12\/2-1024x365.png\"  class=\"wp-image-811\" srcset=\"https:\/\/www.ndnlab.com\/wp-content\/uploads\/2025\/12\/2-1024x365.png 1024w, https:\/\/www.ndnlab.com\/wp-content\/uploads\/2025\/12\/2-300x107.png 300w, https:\/\/www.ndnlab.com\/wp-content\/uploads\/2025\/12\/2-768x274.png 768w, https:\/\/www.ndnlab.com\/wp-content\/uploads\/2025\/12\/2-1536x548.png 1536w, https:\/\/www.ndnlab.com\/wp-content\/uploads\/2025\/12\/2.png 1698w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" title=\"Hattrick: Solving Multi-Class TE using Neural Models\u63d2\u56fe2\" alt=\"Hattrick: Solving Multi-Class TE using Neural Models\u63d2\u56fe2\" \/><\/figure>\n\n\n\n<h2 class=\"wp-block-heading\">\u4e94\u3001\u5b9e\u9a8c\u8bc4\u4f30\uff1aMeta \u751f\u4ea7\u73af\u5883\u7684\u9a8c\u8bc1<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">\u8bba\u6587\u5229\u7528 Meta \u516c\u53f8\u771f\u5b9e\u7684\u5e7f\u57df\u7f51\u62d3\u6251\u548c\u6d41\u91cf\u6570\u636e\u96c6\u5bf9 Hattrick \u8fdb\u884c\u4e86\u4e25\u82db\u6d4b\u8bd5\u3002\u5b9e\u9a8c\u7ed3\u679c\u8868\u660e\uff0c\u5728\u591a\u7ec4\u590d\u6742\u7684\u7f51\u7edc\u62d3\u6251\u4e0b\uff0cHattrick \u8f93\u51fa\u7684\u6d41\u91cf\u5206\u914d\u65b9\u6848\u5728\u6027\u80fd\u4e0a\u9ad8\u5ea6\u903c\u8fd1 LP \u7684\u7406\u8bba\u6700\u4f18\u89e3\uff08Gap \u901a\u5e38\u5728 5% \u4ee5\u5185\uff09\u3002\u66f4\u91cd\u8981\u7684\u662f\uff0c\u5b83\u5728\u786e\u4fdd\u9ad8\u4f18\u5148\u7ea7\u6d41\u91cf\u96f6\u4e22\u5305\u65b9\u9762\uff0c\u8868\u73b0\u8fdc\u4f18\u4e8e\u73b0\u6709\u7684\u5176\u4ed6\u795e\u7ecf\u6a21\u578b\u65b9\u6848\u3002<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">\u6b64\u5916\uff0c\u7814\u7a76\u56e2\u961f\u8fd8\u9a8c\u8bc1\u4e86 Hattrick \u7684<strong>\u62d3\u6251\u6cdb\u5316\u80fd\u529b<\/strong>\u3002\u5f53\u7f51\u7edc\u4e2d\u51fa\u73b0\u94fe\u8def\u65ad\u5f00\u6216\u65b0\u589e\u8282\u70b9\u65f6\uff0c\u9884\u8bad\u7ec3\u597d\u7684 Hattrick \u6a21\u578b\u65e0\u9700\u91cd\u65b0\u8bad\u7ec3\u5373\u53ef\u505a\u51fa\u5408\u7406\u7684\u8c03\u5ea6\u51b3\u7b56\u3002\u5728\u4e0e DOTE\u3001TEAL \u7b49\u4e1a\u754c\u524d\u6cbf\u65b9\u6848\u7684\u5bf9\u6bd4\u4e2d\uff0cHattrick \u5728\u5904\u7406\u591a\u7b49\u7ea7\u4e1a\u52a1\u9700\u6c42\u65f6\u5c55\u73b0\u51fa\u4e86\u538b\u5012\u6027\u7684\u7a33\u5b9a\u6027\u4f18\u52bf\uff0c\u8bc1\u660e\u4e86\u5176\u5728\u5de5\u4e1a\u7ea7\u751f\u4ea7\u73af\u5883\u4e2d\u7684\u90e8\u7f72\u6f5c\u529b\u3002<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"328\" src=\"https:\/\/www.ndnlab.com\/wp-content\/uploads\/2025\/12\/3-1024x328.png\"  class=\"wp-image-812\" srcset=\"https:\/\/www.ndnlab.com\/wp-content\/uploads\/2025\/12\/3-1024x328.png 1024w, https:\/\/www.ndnlab.com\/wp-content\/uploads\/2025\/12\/3-300x96.png 300w, https:\/\/www.ndnlab.com\/wp-content\/uploads\/2025\/12\/3-768x246.png 768w, https:\/\/www.ndnlab.com\/wp-content\/uploads\/2025\/12\/3-1536x492.png 1536w, https:\/\/www.ndnlab.com\/wp-content\/uploads\/2025\/12\/3.png 1779w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" title=\"Hattrick: Solving Multi-Class TE using Neural Models\u63d2\u56fe3\" alt=\"Hattrick: Solving Multi-Class TE using Neural Models\u63d2\u56fe3\" \/><\/figure>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"320\" src=\"https:\/\/www.ndnlab.com\/wp-content\/uploads\/2025\/12\/4-1024x320.png\"  class=\"wp-image-813\" srcset=\"https:\/\/www.ndnlab.com\/wp-content\/uploads\/2025\/12\/4-1024x320.png 1024w, https:\/\/www.ndnlab.com\/wp-content\/uploads\/2025\/12\/4-300x94.png 300w, https:\/\/www.ndnlab.com\/wp-content\/uploads\/2025\/12\/4-768x240.png 768w, https:\/\/www.ndnlab.com\/wp-content\/uploads\/2025\/12\/4-1536x479.png 1536w, https:\/\/www.ndnlab.com\/wp-content\/uploads\/2025\/12\/4.png 1794w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" title=\"Hattrick: Solving Multi-Class TE using Neural Models\u63d2\u56fe4\" alt=\"Hattrick: Solving Multi-Class TE using Neural Models\u63d2\u56fe4\" \/><\/figure>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"972\" height=\"1024\" src=\"https:\/\/www.ndnlab.com\/wp-content\/uploads\/2025\/12\/5-972x1024.png\"  class=\"wp-image-814\" srcset=\"https:\/\/www.ndnlab.com\/wp-content\/uploads\/2025\/12\/5-972x1024.png 972w, https:\/\/www.ndnlab.com\/wp-content\/uploads\/2025\/12\/5-285x300.png 285w, https:\/\/www.ndnlab.com\/wp-content\/uploads\/2025\/12\/5-768x809.png 768w, https:\/\/www.ndnlab.com\/wp-content\/uploads\/2025\/12\/5.png 1095w\" sizes=\"auto, (max-width: 972px) 100vw, 972px\" title=\"Hattrick: Solving Multi-Class TE using Neural Models\u63d2\u56fe5\" alt=\"Hattrick: Solving Multi-Class TE using Neural Models\u63d2\u56fe5\" \/><\/figure>\n\n\n\n<h3 class=\"wp-block-heading\">\u516d\u3001\u603b\u7ed3\u4e0e\u603b\u7ed3\uff1a\u8fc8\u5411\u667a\u80fd\u5316\u7684\u7b97\u7f51\u8c03\u5ea6<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Hattrick<\/strong> \u7684\u51fa\u73b0\u6807\u5fd7\u7740\u6d41\u91cf\u5de5\u7a0b\u7814\u7a76\u4ece\u201c\u901a\u7528\u5316\u4f18\u5316\u201d\u5411\u201c\u7cbe\u7ec6\u5316\u4e1a\u52a1\u611f\u77e5\u201d\u7684\u8de8\u8d8a\u3002\u5b83\u8bc1\u660e\u4e86\u901a\u8fc7\u5de7\u5999\u7684\u67b6\u6784\u8bbe\u8ba1\uff0c\u6df1\u5ea6\u5b66\u4e60\u6a21\u578b\u4e0d\u4ec5\u53ef\u4ee5\u8dd1\u5f97\u5feb\uff0c\u66f4\u53ef\u4ee5\u5904\u7406\u590d\u6742\u7684\u4e1a\u52a1\u4f18\u5148\u7ea7\u7ea6\u675f\u3002\u5bf9\u4e8e\u672a\u6765\u6784\u5efa\u8d85\u5927\u89c4\u6a21\u3001\u5177\u5907\u4e1a\u52a1\u81ea\u9002\u5e94\u80fd\u529b\u7684 AI \u96c6\u7fa4\u7f51\u7edc\u548c\u4e91\u5e7f\u57df\u7f51\uff0cHattrick \u63d0\u4f9b\u4e86\u6781\u5176\u91cd\u8981\u7684\u8bbe\u8ba1\u53c2\u8003\u3002<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">\u5bf9\u4e8e\u60a8\u7684\u7814\u7a76\u65b9\u5411\uff08\u5982 Mamba-NDP \u6846\u67b6\uff09\uff0cHattrick \u5728<strong>\u591a\u4f18\u5148\u7ea7\u5904\u7406<\/strong>\u548c<strong>\u5229\u7528 GNN \u6355\u83b7\u7f51\u7edc\u62d3\u6251\u8bed\u4e49<\/strong>\u65b9\u9762\u7684\u601d\u8def\u5177\u6709\u5f88\u5f3a\u7684\u501f\u9274\u610f\u4e49\u3002\u7279\u522b\u662f\u5b83\u5982\u4f55\u901a\u8fc7\u591a\u9636\u6bb5\u7ed3\u6784\u89e3\u51b3\u201c\u786c\u7ea6\u675f\u201d\u95ee\u9898\u7684\u7ecf\u9a8c\uff0c\u53ef\u4ee5\u76f4\u63a5\u542f\u53d1\u60a8\u5728\u591a\u7ef4\u5ea6\u77e5\u8bc6\u84b8\u998f\u548c MoE \u67b6\u6784\u4e2d\uff0c\u5982\u4f55\u66f4\u597d\u5730\u6743\u8861\u4e0d\u540c\u4efb\u52a1\u6216\u4e13\u5bb6\u4e4b\u95f4\u7684\u4f18\u5148\u7ea7\u3002<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><\/p>\n","protected":false},"excerpt":{"rendered":"<p>SIGCOMM &#8217;25: ACM SIGCOMM 2025 Conference September 8 &#8211; 11, 2025 Coimbra, Portugal Conference Sponsors: SIGCOMM https:\/\/dl.acm.org\/doi\/10.1145\/3718958.3750470 \u4e00\u3001\u7814\u7a76\u80cc\u666f\u4e0e\u52a8\u673a\uff1a\u591a\u4f18\u5148\u7ea7\u6d41\u91cf\u5de5\u7a0b\u7684\u5174\u8d77 \u5728\u73b0\u4ee3\u8d85\u5927\u89c4\u6a21\u5e7f\u57df\u7f51\uff08WAN\uff09\u4e2d\uff0c\u6d41\u91cf\u4e0d\u518d\u88ab\u89c6\u4e3a\u5355\u4e00\u7c7b\u522b\uff0c\u800c\u662f\u6839\u636e\u4e1a\u52a1\u9700\u6c42\u88ab\u5212\u5206\u4e3a\u4e0d\u540c\u7684\u4f18\u5148\u7ea7\uff08 &hellip; <a href=\"https:\/\/www.ndnlab.com\/?p=808\">\u7ee7\u7eed\u9605\u8bfb <span class=\"meta-nav\">&rarr;<\/span><\/a><\/p>\n","protected":false},"author":5,"featured_media":809,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[17,5,23],"tags":[13],"class_list":["post-808","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-17","category-rengongzhineng","category-23","tag-13"],"_links":{"self":[{"href":"https:\/\/www.ndnlab.com\/index.php?rest_route=\/wp\/v2\/posts\/808","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.ndnlab.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.ndnlab.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.ndnlab.com\/index.php?rest_route=\/wp\/v2\/users\/5"}],"replies":[{"embeddable":true,"href":"https:\/\/www.ndnlab.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=808"}],"version-history":[{"count":1,"href":"https:\/\/www.ndnlab.com\/index.php?rest_route=\/wp\/v2\/posts\/808\/revisions"}],"predecessor-version":[{"id":816,"href":"https:\/\/www.ndnlab.com\/index.php?rest_route=\/wp\/v2\/posts\/808\/revisions\/816"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.ndnlab.com\/index.php?rest_route=\/wp\/v2\/media\/809"}],"wp:attachment":[{"href":"https:\/\/www.ndnlab.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=808"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.ndnlab.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=808"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.ndnlab.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=808"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}