#reinforcement_learning

[phonemantra-blog]

A new board of directors is currently being formed Sam Altman will return to the post of CEO of OpenAI after a coup attempt and his removal from this position a few days ago. Former President Greg Brockman, who resigned in protest of Altman's ouster, will also return. The company said it has an "agreement in principle" to bring back Altman along with a new board of directors consisting of Bret Taylor, Larry Summers and Adam D'Angelo. D'Angelo was on the old board of directors that fired Altman on Friday. [caption id="attachment_84151" align="aligncenter" width="750"] OpenAI[/caption] Sam Altman to return as OpenAI CEO after controversial firing All key parties confirmed that they had reached an agreement. Altman said that “everything I have done over the past few days has been aimed at maintaining the unity of this team and its mission.” One of OpenAI's other major investors, Thrive Capital, called Altman's return "the best thing for the company, its employees, those who use its technology, and the world at large." “OpenAI has the potential to become one of the most influential companies in the history of computing,” said Thrive partner Kelly Sims. “Sam and Greg are deeply committed to company integrity and have an unmatched ability to inspire and lead. We are thrilled that they will be returning to the company they founded and helped build into what it is today.” It was previously reported that 96% of OpenAI employees signed a letter saying they could leave and join Sam Altman at Microsoft unless the startup's board resigns and brings back the ousted CEO along with co-founder and former president Greg Brockman.

[jobrxiv]

PhD Researcher University of Innsbruck Are you interested in sustainability? Come join us as a #computerscience PhD to research on a bulding's energy efficiency and optimization! See the full job description on jobRxiv: https://jobrxiv.org/job/university-of-innsbruck-27778-phd-researcher/?feed_id=73097 #building_control #machine_learning #reinforcement_learning #software_engineering #ScienceJobs #hiring #research

[nehaarticles]

Reinforcement Learning Tutorial

Our Reinforcement learning tutorial will give you a complete overview of reinforcement learning, including MDP and Q-learning. In RL tutorial, you will learn the below topics:

[awsexchage]

AWS DeepRacer関連情報まとめ【随時更新】 http://bit.ly/2POQwHz

会社で「AWS DeepRacerリーグで最速伝説つくるぞ！」ってチームに参加することになったので、「AWS DeepRacerってなに？」というところから調べて情報をまとめ始めました。

「この情報おすすめ！」などありましたらコメントいただけると嬉しいです！

公式

AWS DeepRacer – 機械学習を自由自在に操る最速の手段 https://aws.amazon.com/jp/deepracer/ 概要だったり、リーグについてわかります。

AWS DeepRacer とは? – AWS DeepRacer https://docs.aws.amazon.com/ja_jp/deepracer/latest/developerguide/what-is-deepracer.html 公式ドキュメント、日本語化されててありがたいです。「AWS DeepRacerとは」からはじまり、モデルのトレーニング・評価から実車両の走行までひととり揃っています。

AWS-DeepRacer-Getting-Started-Guide.pdf https://d1.awsstatic.com/deepracer/AWS-DeepRacer-Getting-Started-Guide.pdf 車両のセットアップについて写真で解説。（英語)

リーグ情報

DeepRacer リーグの Summit サーキット https://aws.amazon.com/jp/deepracer/summit-circuit/

世界的に開催される 20 の AWS Summit で、モデルの構築、トレーニングを手助けするワークショップに参加しましょう。自宅でトレーニングした自分のモデルを持ち込むこともできます。その後、AWS Summit Expo のトラックで自分のモデルをテストしたり競わせたりできます。20 のイベントの中でいくつでも参加可能です。参加したいレースの Summit に簡単に申し込むことができます。お待ちしています。

AWS Summit Tokyo 2019 | 2019 年 6 月 12 (水) 〜14 (金) 幕張メッセで開催 https://aws.amazon.com/jp/summits/tokyo-2019/

AWS Summit Osaka 2019 | 2019 年 6 月 27 (木) グランフロント大阪で開催 https://aws.amazon.com/jp/summits/osaka-2019/

東京・大阪のDeepRacerワークショップ、受付開始(2019/04/15)後、すぐ満席の模様。。。 oh…

AWS-DeepRacer-League-2019-Official-Rules-English-March-2019.pdf https://d1.awsstatic.com/DeepRacer/AWS-DeepRacer-League-2019-Official-Rules-English-March-2019.pdf リーグの公式ルール(PDF)(英語)

#AWSDeepRacer – Twitter検索 https://twitter.com/search?q=%23AWSDeepRacer&src=typd

シンガポールの最速。はやーい。

Fastest #AWSDeepRacer lap 9.09sec so far… #DRL #AWSSummit #Singapore pic.twitter.com/XTg2CBjEb8 — sunil mallya (@sunilmallya) 2019年4月12日

AWS DeepRacer League https://d3akhm1epsal2g.cloudfront.net/bigscreen/?event=seoul ソウルのリーグ結果(リーダーボード)。

AWS DeepRacer League https://d3akhm1epsal2g.cloudfront.net/bigscreen/?event=singapore シンガポールのリーグ結果(リーダーボード)。

AWS DeepRacer League https://d3akhm1epsal2g.cloudfront.net/bigscreen/?event=paris パリのリーグ結果(リーダーボード)。

Developer at the AWS DeepRacer League Singapore race sets new world record lap time | AWS Machine Learning Blog https://aws.amazon.com/jp/blogs/machine-learning/developer-at-the-aws-deepracer-league-singapore-race-sets-new-world-record-lap-time/ シンガポールの結果。

Announcing the first winner of the AWS DeepRacer League Summit circuit! | AWS Machine Learning Blog https://aws.amazon.com/jp/blogs/machine-learning/announcing-the-first-winner-of-the-aws-deepracer-league-summit-circuit/ サンタクララの結果。

AWS DeepRacer League hits the road for more fun and excitement for developers! | AWS Machine Learning Blog https://aws.amazon.com/jp/blogs/machine-learning/aws-deepracer-league-hits-the-road-for-more-fun-and-excitement-for-developers/ パリの結果。

DeepRacerとは？　系

ごっつ楽しんで簡単に学べる、強化学習超初心者でもDeepRacerで遊んでみよう #reinvent ｜ DevelopersIO https://dev.classmethod.jp/cloud/reinvent-2018-deepracer-getstart/ 利用方法やパラメータの種類について説明してくれています。

AWS DeepRacer について調べてみた | ナレコムAWSレシピ https://recipe.kc-cloud.jp/archives/12261

Developers.IO 2019 at 岡山城でDeepRacerのセッションをしてきました #cmdevio2019 ｜ DevelopersIO https://dev.classmethod.jp/event/developers-io-2019-at-okayama-deepracer/

実践系

AWS DeepRacerで機械学習を始めませんか？ – Speaker Deck https://speakerdeck.com/yamakage92/aws-deepracerdeji-jie-xue-xi-woshi-memasenka

AWSのAIラジコンDeepRacerを開封してみた #reinvent ｜ DevelopersIO https://dev.classmethod.jp/cloud/aws/reinvent2018-unboxing-deepracer/

Amazon SageMakerとAWS RoboMakerでAWS DeepRacerを走らせて学習させるノートブックを試してみた ｜ DevelopersIO https://dev.classmethod.jp/machine-learning/sagemaker-robomaker-deepracer-sample/ AWS DeepRacerのPreview申請がまだ承認されてなくてもモデル作成はできるとのこと。

DeepRacerを走らせる！ー セットアップ編ー – CYDAS Developer’s Blog https://tech.cydas.com/entry/2019/01/29/123838

最初のAWS DeepRacerモデルの作り方 ..というPDFを翻訳してみた – Qiita https://qiita.com/hara2dev/items/3f53bfa8d926087d2c91

Congrats on that thing you won ?!?? https://www.linkedin.com/pulse/congrats-thing-you-won-chris-miller AWS DeepRacerリーグで優勝された方のブログ記事。

DeepRacer Workshopに参加してみました – R3 Cloud Journey https://blog.r3it.com/reinvent2018-oki01-d55e987eb97c

で、サンプルとしてあった中にこんなのがあったので、スタッフに聞いてみると、報酬をどう設定するかという事らしいです。「その辺がインスピレーションとかアイデアとかだよねー（意訳」とか言われました。

ですよねー

ノウハウ

ひとつの AWS アカウント で複数人が AWS DeepRacer を始めた時に注意したこと – Qiita https://qiita.com/dnp-wada/items/18ad01dcc96e6c053cb6

AWS DeepRacer で社内レースをやるために、ラップタイムを計測してみた – Qiita https://qiita.com/dnp-hara/items/cb4081fa5bdfbe749c22

AWS DeepRacerを理論的に考えてみた – Qiita https://qiita.com/dnp-kushihara/items/eda6550ba2a53c7b5d57

DeepRacerの報酬関数トレースログをLogsInsightsでパース – Qiita https://qiita.com/cof123/items/7fdad25d012fc3ac1f4d

勉強会

DeepRacer同好会 第一回オフライン勉強会 #AWSDeepRacerJP – connpass https://connpass.com/event/126766/ DeepRacer同好会 会員用Slackワークスペースが用意されてます。

サンプル

aws-samples/aws-deepracer-workshops: DeepRacer workshop content https://github.com/aws-samples/aws-deepracer-workshops

amazon-sagemaker-examples/rl_deepracer_coach_robomaker.ipynb at master · awslabs/amazon-sagemaker-examples https://github.com/awslabs/amazon-sagemaker-examples/blob/master/reinforcement_learning/rl_deepracer_robomaker_coach_gazebo/rl_deepracer_coach_robomaker.ipynb AWS RoboMaker用のノートブック。

記事

アマゾンAWSのAIロボットカー｢DeepRacer｣の衝撃 —— 強化学習でもアマゾンは制覇する：re:Invent 2018 | BUSINESS INSIDER JAPAN https://www.businessinsider.jp/post-180440

AWS DeepRacerとは｜楽しみながら強化学習を攻略しよう – ビジネス on IT https://www.business-on-it.com/2003-deepracer/

Amazon「AWS Summit Tokyo」6月に東京と大阪で開催、強化学習モデルで走る1/18スケールの自走型ロボットカー・レースも開催！ | ロボスタ https://robotstart.info/2019/02/21/aws-summit-2019.html

元記事はこちら

「AWS DeepRacer関連情報まとめ【随時更新】」

May 07, 2019 at 12:00PM

[ankaapmo]

Structured Learning and Imitation Learning - #Ankaa

Structured Learning and Imitation Learning In a classical prediction use case, the predicted output is either a number (for regression) or category (for classification). A set of training data (x, y) where x is the input and y is the labeled output is provided to train a parameterized predictive model. The model is characterized... https://ankaa-pmo.com/structured-learning-and-imitation-learning/ #Artificial_Intelligence #Imitation_Learning #Reinforcement_Learning #Structured_Learning

[jobrxiv]

PhD Researcher University of Innsbruck Are you interested in sustainability? Come join us as a #computerscience PhD to research on a bulding's energy efficiency and optimization! See the full job description on jobRxiv: https://jobrxiv.org/job/university-of-innsbruck-27778-phd-researcher/?feed_id=72411 #building_control #machine_learning #reinforcement_learning #software_eng. #ScienceJobs #hiring #research

[jobrxiv]

Morphogenetic reverse-engineering: learning collective cell behaviors Collège de France The project focuses on developing inverse morphogenetic approaches to collective cellular behaviors, with a particular emphasis on early embryos. See the full job description on jobRxiv: https://jobrxiv.org/job/college-de-france-27778-morphogenetic-reverse-engineering-learning-collective-cell-behaviors/?feed_id=68562 #biophysics #computational_biology #embryology #inverse_problems #machine_learning #reinforcement_learning #ScienceJobs #hiring #research

[ankaapmo]

Deep Learning Drops the Ball and Dribbles - #Ankaa

Deep Learning Drops the Ball and Dribbles Of all the major league sports in the United States, basketball is the most balletic…in my humble opinion. Basketball demonstrates a fluidity of complex full-body motion all the while interactively and iteratively guiding and correcting the trajectory of a ball which repeatedly... https://ankaa-pmo.com/deep-learning-drops-the-ball-and-dribbles/ #Artificial_Intelligence #Deep_Learning #Physics #Reinforcement_Learning #Video_Game_Animation