{"id":2669,"date":"2016-08-28T22:46:16","date_gmt":"2016-08-28T13:46:16","guid":{"rendered":"http:\/\/tech.akat.info\/?p=2669"},"modified":"2016-08-28T22:46:16","modified_gmt":"2016-08-28T13:46:16","slug":"%e7%ac%ac3%e5%9b%9e-machine-learning-15minutes-%e3%81%ab%e5%8f%82%e5%8a%a0%e3%81%97%e3%81%9f","status":"publish","type":"post","link":"https:\/\/tech.akat.info\/?p=2669","title":{"rendered":"\u7b2c3\u56de Machine Learning 15minutes! \u306b\u53c2\u52a0\u3057\u305f"},"content":{"rendered":"

\u7b2c3\u56de Machine Learning 15minutes!<\/a>\u306b\u53c2\u52a0\u3057\u305f<\/p>\n

-Google\u306e\u6a5f\u68b0\u5b66\u7fd2\u30d7\u30ed\u30c0\u30af\u30c8 \u30b0\u30fc\u30b0\u30eb\u682a\u5f0f\u4f1a\u793e- \u4f50\u85e4 \u4e00\u61b2\u3055\u3093<\/h1>\n

A Neural Network Playground <\/a>
\n\u30cb\u30e5\u30fc\u30e9\u30eb\u30cd\u30c3\u30c8\u30ef\u30fc\u30af\u3092\u7406\u89e3\u3059\u308b\u305f\u3081\u306e\u30c4\u30fc\u30eb<\/p>\n

Google\u3067\u306e\u5229\u7528\u30b7\u30fc\u30f3<\/h2>\n

\u30fb\u30b0\u30fc\u30b0\u30eb\u691c\u7d22
\n\u3000\u6628\u5e74\u304b\u3089\u30c7\u30a3\u30fc\u30d7\u30e9\u30fc\u30cb\u30f3\u30b0(RankBrain)\u30d9\u30fc\u30b9\u306b
\n\u30fbGoogle\u30d5\u30a9\u30c8\u306f\u30c7\u30a3\u30fc\u30d7\u30e9\u30fc\u30cb\u30f3\u30b0\u3067\u5206\u985e\u5206\u3051
\n\u30fb\u82f1\u8a9e\u7248Gmail\u306b\u3064\u3044\u3066\u306f\u30e1\u30fc\u30eb\u8fd4\u4fe1\u3092\u30c7\u30a3\u30fc\u30d7\u30e9\u30fc\u30cb\u30f3\u30b0\u3092\u5229\u7528\u3057\u3066\u4e88\u6e2c\u5909\u63db
\n\u300010%\u306f\u4e88\u6e2c\u5909\u63db\u5185\u5bb9\u304c\u5229\u7528\u3055\u308c\u3066\u3044\u308b
\n\u30fb\u30c7\u30a3\u30fc\u30d7\u30e9\u30fc\u30cb\u30f3\u30b0\u3092\u5229\u7528\u3057\u3066\u3001\u30c7\u30fc\u30bf\u30bb\u30f3\u30bf\u30fc\u96fb\u529b\u52b9\u7387\u304c15%\u30aa\u30d5\u306b<\/p>\n

TensorFlow<\/h2>\n

\u30fbGoogle\u30b5\u30fc\u30d3\u30b9\u3067\u5b9f\u969b\u306b\u5229\u7528\u3055\u308c\u3066\u3044\u308b(\u65e5\u3005\u958b\u767a\u3055\u308c\u3066\u3044\u308b)
\n\u30fb\u5206\u6563\u74b0\u5883\u3092\u524d\u63d0\u3068\u3057\u3066\u8a2d\u8a08\u3055\u308c\u3066\u3044\u308b
\n\u30fb\u5b66\u7fd2\u7d50\u679c\u3092\u30aa\u30d5\u30e9\u30a4\u30f3\u3084\u30b9\u30de\u30fc\u30c8\u30c7\u30d0\u30a4\u30b9\u3067\u3082\u5229\u7528\u53ef\u80fd
\n\u3000(\u4ed6\u306e\u30d5\u30ec\u30fc\u30e0\u30ef\u30fc\u30af\u3088\u308a\u3082\u512a\u308c\u3066\u3044\u308b\u70b9)
\n\u30fb\u73fe\u72b6\u3044\u304b\u306b\u5206\u6563\u3055\u305b\u308b\u304b\u3067\u5404\u30d9\u30f3\u30c0\u304c\u3057\u306e\u304e\u3092\u524a\u3063\u3066\u3044\u308b<\/p>\n

\u8ab0\u3067\u3082\u5229\u7528\u3067\u304d\u308b\u3088\u3046\u306b<\/h2>\n

\u5c02\u9580\u5bb6\u3067\u306a\u3044\u4eba\u305f\u3061\u304c\u5229\u7528\u3057\u3066\u3044\u308b
\n\u4f8b\u3048\u3070\u3001
\n\u30fb\u30ad\u30e5\u30a6\u30ea\u4ed5\u5206\u3051
\n\u30fb\u304b\u3089\u3042\u3052\u30ed\u30dc\u30c3\u30c8<\/p>\n

\u3088\u308a\u6700\u9069\u5316\u3059\u308b\u305f\u3081\u306b<\/h2>\n

\u30cf\u30fc\u30c9\u9762\u306b\u3064\u3044\u3066\u3082\u5f37\u5316\u3057\u3066\u3044\u308b
\n\u2192CloudML\u3067\u305d\u306e\u5a01\u529b\u3092\u5b9f\u611f\u3067\u304d\u308b\u306f\u305a
\n\u2192\u89e3\u6790\u306b\u4f55\u65e5\u304b\u304b\u304b\u308b\u3088\u3046\u3067\u306f\u30c0\u30e1
\n\u30fbJupiter\u30cd\u30c3\u30c8\u30ef\u30fc\u30af
\n\u3000\u30b5\u30fc\u30d0\u9593\u30cd\u30c3\u30c8\u30ef\u30fc\u30af\u306e\u30ec\u30a4\u30c6\u30f3\u30b7\u3092\u4e0b\u3052\u308b\u305f\u3081\u306b\u81ea\u793e\u3067\u958b\u767a
\n\u30fbTensorProcessingUnit
\n\u3000CPU\u3067\u3082\u306a\u304fGPU\u3067\u3082\u306a\u3044
\n\u300010\u500d\u306e\u96fb\u529b\u6027\u80fd\u6bd4
\n\u3000\u2192\u5229\u7528\u3057\u30661\u5e74\u4ee5\u4e0a\u7d4c\u3063\u3066\u3044\u308b\u67af\u308c\u305f\u6280\u8853<\/p>\n

-IBM Watson\u306e\u6a5f\u68b0\u5b66\u7fd2\u30b5\u30fc\u30d3\u30b9\u3068Watson Knowledge Studio- \u65e5\u672c\u30a2\u30a4\u30fb\u30d3\u30fc\u30fb\u30a8\u30e0\u682a\u5f0f\u4f1a\u793e \u5e73\u5c71 \u6bc5\u3055\u3093<\/h1>\n

IBM Watoson\u2192\u30af\u30e9\u30a6\u30c9\u63d0\u4f9b\u3001\u30a8\u30f3\u30bf\u30fc\u30d7\u30e9\u30a4\u30ba\u3092\u610f\u8b58(\u30d7\u30ed\u30c0\u30af\u30c8\u306e\u67a0\u7d44\u307f)
\n\u30b3\u30b0\u30cb\u30c6\u30a3\u30d6\u30b3\u30f3\u30d4\u30e5\u30fc\u30c6\u30a3\u30f3\u30b0\uff1b\u74b0\u5883\u304b\u3089\u3059\u3079\u3066\u7528\u610f(\u8ecd\u4e8b\u95a2\u9023\u3082\u3042\u308b\u305f\u3081\u30aa\u30f3\u30d7\u30ec\u30df\u30b9\u3068\u3044\u3046\u3053\u3068\u3082)<\/p>\n

TensorFlow\u2192\u30aa\u30fc\u30d7\u30f3\u30bd\u30fc\u30b9\u63d0\u4f9b (\u8ab0\u3067\u3082)\u4f7f\u3048\u308b\u3088\u3046\u306b<\/p>\n

IBM Watoson Knowledge Studio
\n\u696d\u754c\u3084\u5206\u91ce\u3054\u3068<\/strong>\u306e\u77e5\u8b58\u306b\u57fa\u3065\u3044\u3066\u3001
\n\u5404\u5206\u91ce\u306e\u8a00\u8449\u306e\u4f7f\u308f\u308c\u65b9\u306e\u5fae\u5999\u306a\u9055\u3044\u307e\u3067Watson\u306b\u6559\u3048\u308b\u3053\u3068\u3092\u53ef\u80fd\u3068\u3059\u308b\u30c4\u30fc\u30eb<\/p>\n

-Dive into Deep Learning Framework- \u682a\u5f0f\u4f1a\u793eABEJA \u7dd2\u65b9 \u8cb4\u7d00<\/h1>\n

\u682a\u5f0f\u4f1a\u793eABEJA IoT, BigData, AI\u3092\u6271\u3063\u3066\u3044\u308b\u4f1a\u793e
\n4\u5e74\u9593\u30c7\u30a3\u30fc\u30d7\u30e9\u30fc\u30cb\u30f3\u30b0\u306b\u5f93\u4e8b\u3057\u3066\u3044\u308b\u3002\u305d\u306e\u969b\u306b\u5229\u7528\u3057\u305f\u30d5\u30ec\u30fc\u30e0\u30ef\u30fc\u30af\u306b\u3064\u3044\u3066\u500b\u4eba\u7684\u611f\u60f3\u3092<\/p>\n

TensorFlow<\/h2>\n

\u30fb\u884c\u5217\u8a08\u7b97\u30e9\u30a4\u30d6\u30e9\u30ea(DeepLearning\u3060\u3051\u3067\u306a\u304f\u3055\u307e\u3056\u307e\u306a\u7528\u9014\u3042\u308a)
\n\u30fbML\u30d7\u30e9\u30c3\u30c8\u30d5\u30a9\u30fc\u30e0\u3092\u63d0\u4f9b\u958b\u59cb\u3059\u308b\u306e\u3067\u3001GPU\u30de\u30b7\u30f3\u304c\u5fc5\u8981\u306a\u304f\u306a\u308b\u304b\u3082
\n\u30fb\u4ed5\u7d44\u307f\u304c\u8907\u96d1\u306e\u305f\u3081\u52c9\u5f37\u3059\u308b\u5fc5\u8981\u304c\u3042\u308b
\n\u30fb\u30b0\u30fc\u30b0\u30eb\u306e\u30d7\u30e9\u30c3\u30c8\u30d5\u30a9\u30fc\u30e0\u3067\u6700\u9069\u5316\u3055\u308c\u3066\u3044\u308b\u305f\u3081\u3001\u74b0\u5883\u306b\u3088\u3063\u3066\u306f\u51e6\u7406\u304c\u91cd\u3044<\/p>\n

K \u30b1\u30e9\u30b9<\/h2>\n

\u30fb\u30b7\u30f3\u30d7\u30eb\u306b\u76f4\u611f\u7684\u306b\u5229\u7528\u3067\u304d\u308b
\n\u30fbTensoleFlow\u306a\u3069\u3092\u30d0\u30c3\u30af\u30a8\u30f3\u30c9\u3068\u3057\u3066\u5229\u7528\u3057\u3066\u3044\u308b
\n\u30fbMulti-GPU\u975e\u5bfe\u5fdc
\n\u30fb\u3068\u308a\u3042\u3048\u305a\u4f7f\u3044\u305f\u3044\u4eba\u5411\u3051<\/p>\n

Chainer<\/h2>\n

\u30fb\u30cd\u30c3\u30c8\u30ef\u30fc\u30af\u8a18\u8f09\u6642\u306e\u67d4\u8edf\u6027\u304c\u9ad8\u3044
\n\u30fb\u5185\u90e8\u304c\u3069\u3046\u306a\u3063\u3066\u3044\u308b\u304b\u308f\u304b\u308a\u3084\u3059\u3044<\/p>\n

-\u5bfe\u8a71bot\u306e\u6280\u8853- \u682a\u5f0f\u4f1a\u793e\u30a6\u30b5\u30ae\u30a3 \u4e94\u6728\u7530 \u548c\u4e5f\u3055\u3093<\/h1>\n

\u81ea\u7136\u306a\u611f\u3058\u3067\u96d1\u8ac7\u3067\u304d\u308b\u305f\u3081\u306bLSTM\u3092\u5229\u7528\u3057\u3066\u3044\u308b
\n\u3000LSTM\u2192Recurrent Neural Network \u3092\u6539\u826f\u3057\u3001\u9577\u671f\u306e\u4f9d\u5b58\u95a2\u4fc2\u3092\u6271\u3048\u308b\u3088\u3046\u306b\u3057\u3066\u3044\u308b<\/p>\n","protected":false},"excerpt":{"rendered":"

\u7b2c3\u56de Machine Learning 15minutes!\u306b\u53c2\u52a0\u3057\u305f -Google\u306e\u6a5f\u68b0\u5b66\u7fd2\u30d7\u30ed\u30c0\u30af\u30c8 \u30b0\u30fc\u30b0\u30eb\u682a\u5f0f\u4f1a\u793e- \u4f50\u85e4 \u4e00\u61b2\u3055\u3093 A Neural Network Playground \u30cb\u30e5\u30fc\u30e9\u30eb\u30cd\u30c3\u30c8\u30ef […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_monsterinsights_skip_tracking":false,"_monsterinsights_sitenote_active":false,"_monsterinsights_sitenote_note":"","_monsterinsights_sitenote_category":0,"footnotes":""},"categories":[94,35],"tags":[],"_links":{"self":[{"href":"https:\/\/tech.akat.info\/index.php?rest_route=\/wp\/v2\/posts\/2669"}],"collection":[{"href":"https:\/\/tech.akat.info\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/tech.akat.info\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/tech.akat.info\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/tech.akat.info\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=2669"}],"version-history":[{"count":1,"href":"https:\/\/tech.akat.info\/index.php?rest_route=\/wp\/v2\/posts\/2669\/revisions"}],"predecessor-version":[{"id":2670,"href":"https:\/\/tech.akat.info\/index.php?rest_route=\/wp\/v2\/posts\/2669\/revisions\/2670"}],"wp:attachment":[{"href":"https:\/\/tech.akat.info\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=2669"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/tech.akat.info\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=2669"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/tech.akat.info\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=2669"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}