{"author_name":"Gasyou","blog_url":"https://gasyou.hatenablog.jp/","image_url":null,"published":"2018-12-07 20:22:47","version":"1.0","width":"100%","categories":["\u958b\u767a\u65e5\u8a18"],"provider_name":"Hatena Blog","height":"190","type":"rich","url":"https://gasyou.hatenablog.jp/entry/2018/12/07/202247","provider_url":"https://hatena.blog","title":"PGLeaf Drei\u5168\u9762\u7684\u306b\u66f8\u304d\u76f4\u3059\u4e8b\u306b\u3057\u305f","description":"\u73fe\u72b6 Policy Gradient Methods for Reinforcement Learning with Function Approximation \u307e\u305a\u3001\u4e0a\u8a18\u8ad6\u6587\u306e\u95a2\u6570\u8fd1\u4f3c\u5668\u306e\u5c0e\u5165\u306f\u65ad\u5ff5\u3057\u307e\u3057\u305f\u3002\u4f55\u3092\u3069\u3046\u3044\u3058\u3063\u3066\u3082\u3001\u7d20\u306ePGLeaf\u4ee5\u4e0a\u306e\u6027\u80fd\u306b\u306f\u306a\u3089\u306a\u304b\u3063\u305f\u306e\u3067\u3002 \u3084\u3063\u305f\u4e8b\u306f\u5927\u4f53\u3053\u3093\u306a\u611f\u3058\u3002 \u8ad6\u6587\u306e\u624b\u6cd5\u3092\u305d\u306e\u307e\u307e\u5b9f\u88c5 \u2192 NG \u3068\u5909\u5f62\u3057\u3066\u307f\u305f \u2192 NG \u306e\u4ee3\u308f\u308a\u306b\u8a55\u4fa1\u5024\u3092\u7528\u3044\u308b \u2192 NG \u306e\u4ee3\u308f\u308a\u306b\u8a55\u4fa1\u5024\u304b\u3089\u4e88\u6e2c\u3057\u305f\u5831\u916c\u306e\u671f\u5f85\u5024\u3092\u7528\u3044\u308b \u2192 NG \u8003\u5bdf \u4e0a\u8a18\u8ad6\u6587\u306b\u3088\u308b\u3068\u3001\u307e\u305a\u3068\u8a55\u4fa1\u95a2\u6570\u306e\u5b66\u7fd2\u3092\u4e26\u884c\u3057\u3066\u884c\u3046\u306e\u3067\u3001\u5b66\u7fd2\u521d\u671f\u306e\u304c\u3042\u307e\u308a\u4fe1\u7528\u51fa\u6765\u306a\u3044\u306e\u3067\u306f\uff1f\u3068\u3044\u3046\u7591\u554f\u3002 \u3053\u306e\u70ba\u3001\u5b66\u7fd2\u521d\u671f\u2026","author_url":"https://blog.hatena.ne.jp/Gasyou/","blog_title":"GA\u5c06\uff1f\u958b\u767a\u65e5\u8a18\uff5e\u738b\u7406\u306e\u305d\u306e\u5148\u3078\uff5e","html":""}