投稿時間:2021-09-08 11:38:42 RSSフィード2021-09-08 11:00 分まとめ(44件)

カテゴリー等	サイト名等	記事タイトル・トレンドワード等	リンクURL	頻出ワード・要約等/検索ボリューム	登録日
IT	気になる、記になる…	Microsoft、パーソナライズ可能な新しいニュースフィードサービス｢Microsoft Start｣を発表	https://taisy0.com/2021/09/08/145043.html	microsoft	2021-09-08 01:05:04
IT	InfoQ	Javaニュース総まとめ: Generational GZC、Kotlin 1.5.30、Jakarta EE 10 Release Plan、SpringOne Conference	https://www.infoq.com/jp/news/2021/09/java-news-roundup-aug23-2021/?utm_campaign=infoq_content&utm_source=infoq&utm_medium=feed&utm_term=global	Javaニュース総まとめGenerationalGZC、Kotlin、JakartaEEReleasePlan、SpringOneConference今週年月日のJava総まとめは、比較的静かだ。	2021-09-08 01:33:00
IT	InfoQ	Gitpodがパブリックおよびプライベートリポジトリの一般提供開始を発表	https://www.infoq.com/jp/news/2021/09/gitpod-general-availability/?utm_campaign=infoq_content&utm_source=infoq&utm_medium=feed&utm_term=global	Gitpodがパブリックおよびプライベートリポジトリの一般提供開始を発表Gitpodは先頃、同社のクラウドベース開発環境をすべてのエンジニアに開放すると発表した。	2021-09-08 01:28:00
IT	InfoQ	ReScript - 完全な型指定を備えたJavaScript代替言語	https://www.infoq.com/jp/news/2021/09/rescript-javascript-alternative/?utm_campaign=infoq_content&utm_source=infoq&utm_medium=feed&utm_term=global	eshertranslatedbyhyoshida	2021-09-08 01:27:00
IT	InfoQ	AWSがセキュリティ調査のためのSecurity Analytics Bootstrapを提供	https://www.infoq.com/jp/news/2021/09/aws-security-analytics-bootstrap/?utm_campaign=infoq_content&utm_source=infoq&utm_medium=feed&utm_term=global	losiotranslatedbyhyoshida	2021-09-08 01:25:00
TECH	Engadget Japanese	Dynabookから小型Windows機 dynaEdge DE200 登場。第11世代Core搭載、AIエンジン対応	https://japanese.engadget.com/dynabook-013114696.html	Dynabookから小型Windows機dynaEdgeDE登場。	2021-09-08 01:31:14
TECH	Engadget Japanese	国内版Galaxy Z Flip3発表、ドコモ・KDDIが10月上旬発売　Fold3と同じくFeliCa対応	https://japanese.engadget.com/flip3-011652571.html	felica	2021-09-08 01:16:52
TECH	Engadget Japanese	速報：国内版 Galaxy Z Fold3 はFeliCa対応、ドコモ・KDDIが10月上旬発売	https://japanese.engadget.com/galaxy-z-fold3-010304976.html	felica	2021-09-08 01:03:04
TECH	Engadget Japanese	『マトリックス・レザレクションズ』ティーザーサイト公開。「赤と青」選んで異なる世界をチラ見せ	https://japanese.engadget.com/the-matrix-resurrections-teaser-010039010.html	映画	2021-09-08 01:00:39
IT	ITmedia 総合記事一覧	[ITmedia News] Twitter、画像や動画を横幅いっぱいに表示するテスト　「違和感ある」「イラストが見やすい」など賛否	https://www.itmedia.co.jp/news/articles/2109/08/news093.html	itmedianewstwitter	2021-09-08 10:49:00
IT	ITmedia 総合記事一覧	[ITmedia News] AWSの大規模障害、原因はネットワークデバイス　新プロトコル処理に潜在的なバグ	https://www.itmedia.co.jp/news/articles/2109/08/news091.html	directconnect	2021-09-08 10:44:00
IT	ITmedia 総合記事一覧	[ITmedia エンタープライズ] 「3年で日本語の自然言語処理がどこまで進化したかを知ってほしい」ELYZA DIGEST開発企業に聞く	https://www.itmedia.co.jp/enterprise/articles/2109/07/news050.html	elyzadigest	2021-09-08 10:30:00
IT	ITmedia 総合記事一覧	[ITmedia News] Ford、AppleとTeslaで自動運転に従事したダグ・フィールド氏を引き抜き	https://www.itmedia.co.jp/news/articles/2109/08/news088.html	apple	2021-09-08 10:29:00
IT	ITmedia 総合記事一覧	[ITmedia News] 山手線「省エネ運転」成功　走行データ分析→加速を減らし、惰行を増やす	https://www.itmedia.co.jp/news/articles/2109/08/news087.html	itmedia	2021-09-08 10:25:00
IT	ITmedia 総合記事一覧	[ITmedia Mobile] 折りたたみスマホ「Galaxy Z Fold3 5G」「Galaxy Z Flip 5G」が10月上旬に日本上陸　ドコモとauが取り扱い	https://www.itmedia.co.jp/mobile/articles/2109/08/news086.html	aukddi	2021-09-08 10:15:00
TECH	Techable（テッカブル）	XRプラットフォーム「STYLY」がVIVEPORTに登場、3作品の個別アプリも	https://techable.jp/archives/161685	psychicvrlab	2021-09-08 01:00:32
AWS	AWS Japan Blog	AWS Cloud9 が大阪リージョンでご利用可能になりました	https://aws.amazon.com/jp/blogs/news/aws-cloud9-osaka/	私自身も参加しハンズオン等を担当させていただきましたが、学習と遊びが融合したとても熱い、クラウドに触れる日でした。	2021-09-08 01:25:01
Program	[全てのタグ]の新着質問一覧｜teratail（テラテイル）	JavaScriptの非同期処理の呼び出し方	https://teratail.com/questions/358314?rss=all	JavaScriptの非同期処理の呼び出し方JavaScriptで非同期処理を呼び出す際にawaitって必要なのでしょうか。	2021-09-08 10:59:17
Program	[全てのタグ]の新着質問一覧｜teratail（テラテイル）	Windows起動時のパスワード入力ってハッキング不可能なんじゃないんですか？	https://teratail.com/questions/358313?rss=all	windows	2021-09-08 10:57:21
Program	[全てのタグ]の新着質問一覧｜teratail（テラテイル）	WordPress ブラウザタブに表示する名前	https://teratail.com/questions/358312?rss=all	WordPressブラウザタブに表示する名前よろしくお願いいたします。	2021-09-08 10:55:54
Program	[全てのタグ]の新着質問一覧｜teratail（テラテイル）	[WPF] 表示/非表示できる操作画面を作りたい	https://teratail.com/questions/358311?rss=all	WPF表示非表示できる操作画面を作りたい前提・実現したいこと動画プレイヤーを学習の一環として作ってみようと考えていますが、OnOffできる操作パネルのレイアウト方法で困っています。	2021-09-08 10:54:26
Program	[全てのタグ]の新着質問一覧｜teratail（テラテイル）	StarRatingで読み込まれない部分がある.	https://teratail.com/questions/358310?rss=all	starrating	2021-09-08 10:53:14
Program	[全てのタグ]の新着質問一覧｜teratail（テラテイル）	DOMParserはなぜnewしてから使う設計になっているのでしょうか？	https://teratail.com/questions/358309?rss=all	DOMParserはなぜnewしてから使う設計になっているのでしょうかJavaScriptのDOMParserについての質問です。	2021-09-08 10:51:27
Program	[全てのタグ]の新着質問一覧｜teratail（テラテイル）	１～100の乱数を配列にセットするには	https://teratail.com/questions/358308?rss=all		2021-09-08 10:50:01
Program	[全てのタグ]の新着質問一覧｜teratail（テラテイル）	HTMLでPDFの埋め込みをするには？	https://teratail.com/questions/358307?rss=all	HTMLでPDFの埋め込みをするにはこんにちは。	2021-09-08 10:49:15
Program	[全てのタグ]の新着質問一覧｜teratail（テラテイル）	python　__class__.__name__　動作がわからない。	https://teratail.com/questions/358306?rss=all	pythonclassname動作がわからない。	2021-09-08 10:44:45
Program	[全てのタグ]の新着質問一覧｜teratail（テラテイル）	Google Colab での GmailAPI の認証	https://teratail.com/questions/358305?rss=all	GoogleColabでのGmailAPIの認証GooglenbspColabnbspを使用するとGoogleDriveやGooglenbspSpreadsheetはcredentialのjsonがなくても以下を実行することで認証出来ると思います。	2021-09-08 10:25:02
Program	[全てのタグ]の新着質問一覧｜teratail（テラテイル）	Google tag managerでタグを</body>タグの直前に配置した	https://teratail.com/questions/358304?rss=all	googlenbsptagnbspmanager	2021-09-08 10:24:43
Program	[全てのタグ]の新着質問一覧｜teratail（テラテイル）	同じグループに指定しつつ、そのグループ内の特定html要素に限定した指定も行える属性	https://teratail.com/questions/358303?rss=all		2021-09-08 10:22:52
AWS	AWSタグが付けられた新着投稿 - Qiita	作成したEC2にWebサーバー（nginx）を立ち上げるまで。	https://qiita.com/kuma15/items/c6f5938cc4103e65d635	マニュアルの確認manamazonlinuxextras利用可能なリストを表示amazonlinuxextras管理者権限でインストールsudoamazonlinuxextrasinstallnginxnginxを起動sudosystemctlstartnginxstatusnginxのステータスを確認起動されていることを確認systemctlstatusnginxstatusポート番の状態を確認sudolsofiWebサーバーの動作確認ECセキュリティグループのインバウンドルールを編集。	2021-09-08 10:16:10
技術ブログ	Developers.IO	AWS Systems Manager Change Calendarで祝日を設定したジョブ実行が可能になりました	https://dev.classmethod.jp/articles/ssm-change-calendar-import-holiday/	ohumanlaborisnohumanerror	2021-09-08 01:37:16
海外TECH	DEV Community	Vocabulary Building Day 12	https://dev.to/vindhyasadanand/building-vocabulary-day-11-4ib8	Vocabulary Building Day reticent adjective not revealing one s thoughts or feelings readily Checkout Pronunciationlambast verb criticize someone or something harshly Checkout Pronunciationperfidy adjective untrustworthy Checkout Pronunciationmawkish adjective overly sentimental to the point that is disgusting Checkout Pronunciationaesthetic adjective concerned with beauty or the appreciation of beauty Checkout PronunciationI hope this helps you if you are preparing for GRE exams	2021-09-08 01:51:56
海外TECH	DEV Community	Design reinforcement learning agents using Unity ML-Agents	https://dev.to/joooyz/design-reinforcement-learning-agents-using-unity-ml-agents-58f0	Design reinforcement learning agents using Unity ML AgentsThis article is part of the series A hands on introduction to deep reinforcement learning using Unity ML Agents It s also suitable for anyone new to Unity interested in using ML Agents for their own reinforcement learning project Recap and overviewIn part we built a D physics based volleyball environment in Unity We also added rewards to encourage agents to volley In this tutorial we ll add agents to the environment The goal is to let them observe and interact with the environment so that we can train them later using deep reinforcement learning Letting our agents make decisionsWe want our agent to learn which actions to take given a certain state of the environment ーe g if the ball is on our side of the court our agent should get it before it hits the floor The goal of reinforcement learning is to learn the best policy a mapping of states to actions that will maximise possible rewards The theory behind how reinforcement learning algorithms achieve this is beyond the scope of this series but the courses I shared in the series introduction will cover it in great depth While training the agent will either take actions At random to explore which actions lead to rewards and which don t From its current policy the optimal action given the current state ML Agents provides a convenient Decision Requester component which will handle the alternation between these for us during training To add a Decision Requester Select the PurpleAgent game object within the PurplePlayArea parent Add Component gt Decision Requester Leave decision period default as Defining the agent behaviorBoth agents are already set up with the VolleyballAgent cs script and Behavior Parameters component which we ll come back to later In this part we ll walk through VolleyballAgent cs This script contains all the logic that defines the agents actions and observations It contains some helper methods already Start ーcalled when the environment is first rendered Grabs the parent Volleyball environment and saves it to a variable envController for easy reference to its methods later Initialize ーcalled when the agent is first initialized Grabs some useful constants and objects Also sets agentRot to ensure symmetry so that the same policy can be shared between both agents MoveTowards CheckIfGrounded amp Jump ーfrom ML Agents sample projects Used for jumping OnCollisionEnter ーcalled when the Agent collides with something Used to update lastHitter to decide which agent gets penalized if the ball is hit out of bounds or rewarded if hit over the net Adding an agent in Unity ML Agents usually involves extending the base Agent class and implementing the following methods OnActionReceived Heuristic CollectObservations OnEpisodeBegin Note usually used for resetting starting conditions We don t implement it here because the reset logic is already defined at the environment level in VolleyballEnvController This makes more sense for us since we also need to reset the ball and not just the agents Agent actionsAt a high level the Decision Requester will select an action for our agent to take and trigger OnActionReceived This in turn calls MoveAgent MoveAgent This method resolves the selected action Within the MoveAgent method start by declaring vector variables for our agents direction and rotation movements var dirToGo   Vector zero var rotateDir   Vector zero We ll also add a grounded check to see whether its possible for the agent to jump var grounded CheckIfGrounded The actions passed into this method actionBuffers DiscreteActions will be an array of integers which we ll map to some behavior It s not important which order we assign them as long as they remain consistent var dirToGoForwardAction act var rotateDirAction act var dirToGoSideAction act var jumpAction act In Unity every object has a transform class that stores its position rotation and scale We ll use it to create a vector pointing to the correct direction in which we want our agent to move Based on the previous assignment order this is how we ll map our actions to behaviors dirToGoForwardAction Do nothing Move forward Move backward rotateDirAction Do nothing Rotate clockwise Rotate anti clockwise dirToGoSideAction Do nothing Move left Move right jumpAction Don t jump Jump Add to the MoveAgent method if dirToGoForwardAction dirToGo grounded f f transform forward f else if dirToGoForwardAction dirToGo grounded f f transform forward volleyballSettings speedReductionFactor f if rotateDirAction rotateDir transform up f else if rotateDirAction rotateDir transform up f if dirToGoSideAction dirToGo grounded f f transform right volleyballSettings speedReductionFactor f else if dirToGoSideAction dirToGo grounded f f transform right volleyballSettings speedReductionFactor if jumpAction if jumpingTime lt f amp amp grounded Jump Note volleyballSettings speedReductionFactor is a constant that slows backwards and strafe movement to be more realistic Next apply the movement using Unity s provided Rotate and AddForce methods transform Rotate rotateDir Time fixedDeltaTime f agentRb AddForce agentRot dirToGo volleyballSettings agentRunSpeed ForceMode VelocityChange Finally add in the logic for controlling jump behavior makes the agent physically jump if jumpingTime gt f jumpTargetPos new Vector agentRb position x jumpStartingPos y volleyballSettings agentJumpHeight agentRb position z agentRot dirToGo MoveTowards jumpTargetPos agentRb volleyballSettings agentJumpVelocity volleyballSettings agentJumpVelocityMaxChange provides a downward force to end the jumpif jumpingTime gt f amp amp grounded agentRb AddForce Vector down volleyballSettings fallingForce ForceMode Acceleration controls the jump sequenceif jumpingTime gt f jumpingTime Time fixedDeltaTime Heuristic To test that we ve resolved the actions properly lets implement the Heuristic method This will map actions to a keyboard input so that we can playtest as a human controller Add to Heuristic var discreteActionsOut actionsOut DiscreteActions if Input GetKey KeyCode D rotate right discreteActionsOut if Input GetKey KeyCode W Input GetKey KeyCode UpArrow forward discreteActionsOut if Input GetKey KeyCode A rotate left discreteActionsOut if Input GetKey KeyCode S Input GetKey KeyCode DownArrow backward discreteActionsOut if Input GetKey KeyCode LeftArrow move left discreteActionsOut if Input GetKey KeyCode RightArrow move right discreteActionsOut discreteActionsOut Input GetKey KeyCode Space Save your script and return to the Unity editor In the Behavior Parameters component of the PurpleAgent Set Behavior Type to Heuristic Only This will call the Heuristic method Set up the Actions Discrete Branches Branch Size No movement move forward move backward Branch Size No movement move left move right Branch Size No rotation rotate clockwise rotate anti clockwise Branch Size No Jump jump Press ️in the editor and you ll be able to use the arrow keys or WASD and space bar to control your agent Note It might be easier to playtest if you comment out the EndEpisode calls in ResolveEvent of VolleyballEnvController cs to stop the episode resetting ObservationsObservations are how our agent sees its environment In ML Agents there are types of observations we can use Vectors ー direct information about our environment e g a list of floats containing the position scale velocity etc of objects Raycasts ー beams that shoot out from the agent and detect nearby objectsVisual camera inputIn this project we ll implement vector observations to keep things simple The goal is to include only the observations that are relevant for making an informed decision With some trial and error here s what I decided to use for observations Agent s y rotation float Agent s x y z velocity floats Agent s x y z normalized vector to the ball i e direction to the ball floats Ball s x y z velocity floats This is a total of vector observations Feel free to experiment with different observations For example you might ve noticed that the agent knows nothing about its opponent This ends up working fine for training a simple agent that can bounce the ball over the net but won t be great at training a competitive agent that wants to win Also note that selecting observations depends on your goal If you re trying to replicate a real world scenario these observations won t make sense It would be very unlikely for a player to know these direct values about the environment To add observations you ll need to implement the Agent class CollectObservations method public override void CollectObservations VectorSensor sensor Agent rotation float sensor AddObservation this transform rotation y Vector from agent to ball direction to ball floats Vector toBall new Vector ballRb transform position x this transform position x agentRot ballRb transform position y this transform position y ballRb transform position z this transform position z agentRot sensor AddObservation toBall normalized Distance from the ball float sensor AddObservation toBall magnitude Agent velocity floats sensor AddObservation agentRb velocity Ball velocity floats sensor AddObservation ballRb velocity y sensor AddObservation ballRb velocity z agentRot sensor AddObservation ballRb velocity x agentRot Now we ll finish setting up the Behavior Parameters Set Behavior Name to Volleyball Later this is how our trainer will know which agent to train Set Vector Observation Space Size Stacked Vectors Wrap upYou re now all set up to train your reinforcement learning agents If you get stuck check out the pre configured BlueAgent or see the full source code in the Ultimate Volleyball project repo In the next section we ll train our agents using PPO ーa state of the art RL algorithm provided out of the box by Unity ML Agents Part Training agents using PPO Coming soon If you have any feedback or questions please let me know	2021-09-08 01:22:22
Apple	AppleInsider - Frontpage News	Apple's iPad stock is running low amid rumors of refresh	https://appleinsider.com/articles/21/09/08/apples-ipad-stock-is-running-low-amid-rumors-of-refresh?utm_medium=rss	Apple x s iPad stock is running low amid rumors of refreshStock of Apple s most affordable iPad model is running low at retail stores and online suggesting a rumored refresh of the entry level tablet is near Apple com has severely limited iPad availability with nearly all configurations showing significant delays in delivery time Spot checks conducted by AppleInsider confirm the device is in low supply across the U S and many major markets including China Japan and the UK In the U S shipments for Space Gray and gold iPad variants are seeing delivery dates pushed back to between Oct and Oct The delay applies to all storage and connectivity configurations save for the GB Wi Fi Cellular model in Space Gray which ships between Sept and Oct in the U S Read more	2021-09-08 01:51:21
ニュース	ジェトロ ビジネスニュース（通商弘報）	社会的距離確保の第4段階を10月3日まで延長	https://www.jetro.go.jp/biznews/2021/09/c3cb74606911235c.html	距離	2021-09-08 01:30:00
ニュース	BBC News - Home	MPs to vote on health and social care tax rise	https://www.bbc.co.uk/news/uk-politics-58483036?at_medium=RSS&at_campaign=KARANGA	percentage	2021-09-08 01:02:26
ニュース	BBC News - Home	Teenager Fernandez stuns fifth seed Svitolina to reach US Open semis	https://www.bbc.co.uk/sport/tennis/58482860?at_medium=RSS&at_campaign=KARANGA	Teenager Fernandez stuns fifth seed Svitolina to reach US Open semisTeenager Leylah Fernandez produces another fearless performance to beat Elina Svitolina and set up a US Open semi final with Aryna Sabalenka	2021-09-08 01:04:16
LifeHuck	ライフハッカー［日本版］	【Amazonタイムセール中！】1,839円のディアドラのスニーカーや799円のランニングポーチなど	https://www.lifehacker.jp/2021/09/amazon-timesale-0908-1.html	amazon	2021-09-08 10:30:00
北海道	北海道新聞	東証、４７円高　午前９時１５分現在	https://www.hokkaido-np.co.jp/article/586880/	日経平均株価	2021-09-08 10:05:00
ビジネス	東洋経済オンライン	人の限界を超えるテクノロジーは｢神への冒涜｣か ネオ･ヒューマンと｢人工流れ星｣で考える科学 | 読書 | 東洋経済オンライン	https://toyokeizai.net/articles/-/452551?utm_source=rss&utm_medium=http&utm_campaign=link_back	東洋経済オンライン	2021-09-08 11:00:00
ニュース	Newsweek	「ヒュブリス」だった東京五輪が日本に残す教訓	https://www.newsweekjapan.jp/stories/world/2021/09/post-97060.php	組織委員会やIOC、政府の誰が何を担当するかが明確にされてないことも大きな問題だ。	2021-09-08 10:30:00
IT	週刊アスキー	マクドナルドの最大1000円お得なデリバリーメニュー「おもいっきりポテナゲセット」を頼んでみた	https://weekly.ascii.jp/elem/000/004/068/4068396/	山口真央	2021-09-08 10:30:00
IT	週刊アスキー	ドコモとauから「Galaxy Z Fold3 5G」「Galaxy Z Flip3 5G」発売　ともにおサイフケータイ対応	https://weekly.ascii.jp/elem/000/004/068/4068526/	galaxyzflipg	2021-09-08 10:30:00
マーケティング	AdverTimes	手作業から生まれた数々の作品、 アートディレクター葛西薫個展「NOSTALGIA」	https://www.advertimes.com/20210908/article362618/		2021-09-08 01:52:30