{"published":"2021-10-27 22:32:49","width":"100%","blog_url":"https://randommemory.hatenablog.com/","author_name":"derwind","description":"Perplexity \u306f\u3061\u3087\u3063\u3068\u5f0f\u304c\u308f\u304b\u308a\u306b\u304f\u3044\u3051\u3069\\begin{align*} L = - \\frac{1}{N} \\sum_n \\sum_k t_{nk} \\log y_{nk} \\end{align*}\u306b\u304a\u3044\u3066\u3001$N=1$ \u3068\u601d\u3063\u3066\u3001$y_{nk} = p$ \u3068\u601d\u3046\u3068 $L= -\\log p$ \u306b\u306a\u3063\u3066\u3001 $e^L = e^{-\\log p} = 1/p$ \u3092\u5f97\u308b\u306e\u3067\u300c\u78ba\u7387\u306e\u9006\u6570\u300d\u307f\u305f\u3044\u306a\u611f\u3058\u306b\u306a\u3063\u3066\u3044\u308b\u3002\u4e00\u822c\u306e\u30b1\u30fc\u30b9\u3067\u3082 $- \\frac{1}{N} \\sum_n \\sum_k t_{nk} \\log y_{nk} = - \\frac{1}{N}(\\log y_{1,k_1} + \\c\u2026","categories":["machine_learning"],"provider_url":"https://hatena.blog","blog_title":"\u3089\u3093\u3060\u3080\u306a\u8a18\u61b6","image_url":null,"html":"<iframe src=\"https://hatenablog-parts.com/embed?url=https%3A%2F%2Frandommemory.hatenablog.com%2Fentry%2F2021%2F10%2F27%2F223249\" title=\"\u30bc\u30ed\u3064\u304f 2 (7) - \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>","version":"1.0","author_url":"https://blog.hatena.ne.jp/derwind/","url":"https://randommemory.hatenablog.com/entry/2021/10/27/223249","height":"190","provider_name":"Hatena Blog","title":"\u30bc\u30ed\u3064\u304f 2 (7)","type":"rich"}