{"blog_title":"DeepLearning\u3092\u52c9\u5f37\u3059\u308b\u4eba","title":"Prioritized Experience Replay","image_url":"http://chart.apis.google.com/chart?cht=tx&chl=%5Cdelta","type":"rich","html":"<iframe src=\"https://hatenablog-parts.com/embed?url=https%3A%2F%2Fwanwannodao.hatenablog.com%2Fentry%2F2017%2F03%2F31%2F194646\" title=\"Prioritized Experience Replay - DeepLearning\u3092\u52c9\u5f37\u3059\u308b\u4eba\" class=\"embed-card embed-blogcard\" scrolling=\"no\" frameborder=\"0\" style=\"display: block; width: 100%; height: 190px; max-width: 500px; margin: 10px 0px;\"></iframe>","provider_url":"https://hatena.blog","height":"190","provider_name":"Hatena Blog","blog_url":"https://wanwannodao.hatenablog.com/","description":"[1511.05952] Prioritized Experience Replay \u8ad6\u6587\u307e\u3068\u3081 Online RL\u306e\u554f\u984c\u70b9 \u9077\u79fb(transition)\u9593\u306e\u4f9d\u5b58\u95a2\u4fc2\u306e\u5f71\u97ff\u304c\u5927\u304d\u3044 \u30ec\u30a2\u306a\u9077\u79fb\u3092\u3059\u3050\u306b\u6368\u3066\u3066\u3057\u307e\u3046 \u305d\u3053\u3067\u3001 Experience Replay(ER) DQN\u3067\u306f\u3001replay mem.\u304b\u3089\u30e9\u30f3\u30c0\u30e0\u30b5\u30f3\u30d7\u30ea\u30f3\u30b0\u3057\u305f\u30df\u30cb\u30d0\u30c3\u30c1\u3092\u4f7f\u3063\u3066\u66f4\u65b0\u3059\u308b. ER\u3092\u3088\u308a\u52b9\u7387\u7684\u3001\u52b9\u679c\u7684\u306b\u3059\u308b\u306b\u306f\uff1f\uff1f \u512a\u5148\u9806\u4f4d\u3092\u4ed8\u3051\u3066\u30b5\u30f3\u30d7\u30ea\u30f3\u30b0\u3059\u308b. \u3088\u308a\u5b66\u7fd2\u306e\u9032\u884c\u304c\u671f\u5f85\u3055\u308c\u308b\u9077\u79fb\u3092\u3088\u308a\u983b\u7e41\u306b\u30ea\u30d7\u30ec\u30a4\u3059\u308b\u3088\u3046\u306b\u3059\u308b. \u30cb\u30e5\u30fc\u30ed\u30b5\u30a4\u30a8\u30f3\u30b9\u7684\u306a\u8a71 \u4f11\u61a9\u3057\u3066\u3044\u308b\u6642\u3084\u5bdd\u3066\u3044\u308b\u6642\u306b\u3001\u6d77\u99ac\u306f\u904e\u53bb\u306e\u7d4c\u9a13\u3092\"\u30ea\u30d7\u30ec\u30a4\"\u3057\u3066\u3044\u308b.\u2026","version":"1.0","author_name":"wanwannodao","published":"2017-03-31 19:46:46","width":"100%","author_url":"https://blog.hatena.ne.jp/wanwannodao/","url":"https://wanwannodao.hatenablog.com/entry/2017/03/31/194646","categories":["DeepLearning","Python","ReinforcementLearning","TensorFlow","\u5f37\u5316\u5b66\u7fd2","\u81ea\u5206\u7528"]}