{"categories":["machine_learning"],"type":"rich","published":"2019-12-07 21:47:19","author_name":"derwind","url":"https://randommemory.hatenablog.com/entry/2019/12/07/214719","blog_url":"https://randommemory.hatenablog.com/","html":"<iframe src=\"https://hatenablog-parts.com/embed?url=https%3A%2F%2Frandommemory.hatenablog.com%2Fentry%2F2019%2F12%2F07%2F214719\" title=\"PRML\u3068\u30cb\u30e5\u30fc\u30e9\u30eb\u30cd\u30c3\u30c8\u30ef\u30fc\u30af - \u3089\u3093\u3060\u3080\u306a\u8a18\u61b6\" class=\"embed-card embed-blogcard\" scrolling=\"no\" frameborder=\"0\" style=\"display: block; width: 100%; height: 190px; max-width: 500px; margin: 10px 0px;\"></iframe>","height":"190","title":"PRML\u3068\u30cb\u30e5\u30fc\u30e9\u30eb\u30cd\u30c3\u30c8\u30ef\u30fc\u30af","description":"Pattern Recognition and Machine Learning - Microsoft Research 2006\u5e74\u306e\u672c\u3060\u304b\u3089\u7d50\u69cb\u53e4\u3044\u3093\u304b\u306a\u3068\u601d\u3044\u304d\u3084\u3001\u30cb\u30e5\u30fc\u30e9\u30eb\u30cd\u30c3\u30c8\u30ef\u30fc\u30af\u306e\u3068\u3053\u308d\u3067\u306f\u666e\u901a\u306b activation function \u3068\u3057\u3066 sigmoid \u3060\u3068\u304b tanh \u3068\u304b\u51fa\u3066\u304f\u308b\u3057\u3001\u5206\u985e\u554f\u984c\u306a\u3089 softmax \u3092\u901a\u3059\u7684\u306a\u8a71\u304c\u66f8\u3044\u3066\u3042\u3063\u3066\u3082\u3046\u3053\u306e\u9803\u306b\u306f\u307b\u307c\u307b\u307c\u78ba\u7acb\u3055\u308c\u3068\u3063\u305f\u3093\u304b\u306a\u3068\u601d\u3048\u305f\u3002ReLU \u306f2011\u5e74\u306b\u51fa\u3066\u304d\u305f\u3082\u306e\u307f\u305f\u3044\u3060\u304b\u3089\u6d41\u77f3\u306b\u8f09\u3063\u3066\u306a\u3044\u3051\u3069\u3002\u6d3b\u6027\u5316\u95a2\u6570 - Wikipedia 2013\u5e74\u306bmax(0.01x, x) \u304cLReL (leaky recti\u2026","width":"100%","provider_name":"Hatena Blog","version":"1.0","image_url":null,"provider_url":"https://hatena.blog","author_url":"https://blog.hatena.ne.jp/derwind/","blog_title":"\u3089\u3093\u3060\u3080\u306a\u8a18\u61b6"}