BEGIN:VCALENDAR VERSION:2.0 PRODID:-//https://techplay.jp//JP CALSCALE:GREGORIAN METHOD:PUBLISH X-WR-CALDESC:CMU-MLD = AIRC Joint Workhop on AI Co-evolving with Humans in the Real-World X-WR-CALNAME:CMU-MLD = AIRC Joint Workhop on AI Co-evolving with Humans in the Real-World X-WR-TIMEZONE:Asia/Tokyo BEGIN:VTIMEZONE TZID:Asia/Tokyo BEGIN:STANDARD DTSTART:19700101T000000 TZOFFSETFROM:+0900 TZOFFSETTO:+0900 TZNAME:JST END:STANDARD END:VTIMEZONE BEGIN:VEVENT UID:701429@techplay.jp SUMMARY:CMU-MLD = AIRC Joint Workhop on AI Co-evolving with Humans in the R eal-World DTSTART;TZID=Asia/Tokyo:20181102T133000 DTEND;TZID=Asia/Tokyo:20181102T170000 DTSTAMP:20260420T231315Z CREATED:20181009T021313Z DESCRIPTION:イベント詳細はこちら\nhttps://techplay.jp/event/70142 9?utm_medium=referral&utm_source=ics&utm_campaign=ics\n\nCMU-MLD = AIRC J oint Workhop on AI Co-evolving with Humans in the Real-World\n\nすべ ての講演は英語で行われます。通訳はございません。\ n\n人工知能研究における国際拠点の構築を目指してい ます産総研は、米国カーネギーメロン大学との連携も すすめています。本ワークショップでは、カーネギー メロン大学からお招きした講師にご講演頂き、合わせ て産総研人工知能研究センターの研究者が講演し、本 連携に関する最新動向をご紹介します。\n\n基本情報\n\ n\n名称：CMU-MLD = AIRC Joint Workhop on\n　　　AI Co-evolving with Humans in the Real-World\n日時：2018-11-02（金）13:30 - 17:00\n受 付時間：13:00 - 15:00\n場所：テレコムセンタービル東棟1 4階 MONO研修室\nURL：https://mono.jpn.com/telecom-center-access/\n 定員：150名\n参加費用：無料\n主催：産業技術総合研究 所人工知能研究センター\n連絡先：人工知能研究戦略 部 問い合わせ窓口\n\n\n注意事項\n\n\n他の方に参加の機 会をお譲りするためにも、参加ができないと分かった 場合は早めのキャンセルをお願いします。\n本名での ご登録をお願いします。\n産総研は、お送りいただい た情報をセミナー運営以外の目的には使用しません。\ n懇親会の予定はありません。\n\n\nプログラム\n\n\n\n\n13 :30 - 14:00\n\n産業技術総合研究所 人工知能研究センター 　麻生 英樹 副研究センター長 \n「Introduction of AIRC\, AIS T\n　- An open innovation hub for AI collaborating with humans in the re al world」\n概要：\nAI\, which has been developing very rapidly mainl y based on machine learning using the big data gathered through services on the internet\, is now being incorporated into various services in the real world and becoming the most important technological infrastructure f or the data-driven smart society.\nAIRC is established in May 2015 to ser ve as an open innovation hub for promoting large-scale AI research in col laboration with researchers from Japan and worldwide.\nSince its establis hment\, AIRC has been focusing on the research toward the AI collaboratin g with humans in the real-world\, where explainability\, interpretability \, and understandability of AI are very important.\nIn this talk\, I woul d like to introduce current activities of AIRC briefly and discuss the fu ture direction of AI research.\n\n\n\n\n\n14:00 - 15:00\n\n米国カー ネギーメロン大学　Manuela Veloso教授（Head of J.P. Morgan AI Research）\n「Towards a Lasting Human-AI Interaction」\n概要：\nArt ificial intelligence\, including extensive data processing\, decision mak ing and execution\, and learning from experience\, offers new challenges for an effective human-AI interaction.\nThis talk delves into multiple ro les humans can have in such interaction\, as well as the underlying chall enges to AI in particular in terms of collaboration and interpretability. \nThe presentation is grounded within the context of autonomous mobile se rvice robots\, and applications to other areas.\n\n経歴：\nManuela M. Veloso has recently joined J.P.Morgan Chase to create and head an Artific ial Intelligence (AI) Research Center. Veloso is on leave from Carnegie M ellon University (CMU) where she is Herbert A. Simon University Professor in the School of Computer Science\, and where she was the Head of the Ma chine Learning Department until June 2018. She researches in AI\, Robotic s\, and Machine Learning. At CMU\, she founded and directs the CORAL rese arch laboratory\, for the study of autonomous agents that Collaborate\, O bserve\, Reason\, Act\, and Learn. Veloso and her students research a var iety of autonomous robots\, including mobile service robots and soccer ro bots. Veloso is AAAI Fellow\, ACM Fellow\, AAAS Fellow\, and IEEE Fellow\ , Einstein Chair Professor of the Chinese National Academy of Science\, t he co-founder and past President of RoboCup\, and past President of AAAI. See www.cs.cmu.edu/~mmv for further information\, including publications. \n【CORAL Group - Carnegie Mellon University】\nWelcome to the CORAL re search group\, led by Professor Manuela Veloso. We research on the scient ific and engineering challenges of creating teams of intelligent agents i n complex\, dynamic\, and uncertain environments\, in particular adversar ial environments\, such as robot soccer.\n\n関連リンク：\nTop A.I. expert to join J.P. Morgan\n\n\n\n\n15:00 - 15:15\n\n休憩\n\n\n\n\n15:1 5 - 15:45\n\n産業技術総合研究所 人工知能研究センター 機械学習研究チーム　椿 真史 研究員\n「Graph Neural Netwo rks for Molecules\n　：Interpretable Applications for Biological and Ma terial Data」\n概要：\nGraph neural networks (GNNs) for molecules hav e a potential to be applied to bioinformatics\, chemoinformatics\, and ma terial informatics.\nFor example\, in bioinformatics\, the prediction of compound-protein interactions plays an important role in the virtual scre ening for drug discovery.\nAs another example\, in material informatics\, the discovery of molecules with specific properties is crucial to develo ping effective materials.\nIn this presentation\, we introdue our recentl y proposed new GNN models for these problems\; in particular\, the models involving some aspects derived from the biological and material knowledg e.\nWe believe that this leads to the interpretable applications for bioi nformatics and material informatics.\n\n関連リンク：\n人工知能 研究センター 機械学習研究チーム \n\n\n\n\n15:45 - 16:45\n\ n米国カーネギーメロン大学　Pradeep Ravikumar准教授\n「Ex plainable Artificial Intelligence via Representer Points」\n概要：\nA s machine learning systems start to be more widely used\, we are starting to care not just about the accuracy and speed\nof predictions\, but also why the ML system made its specific predictions.\nIn the case of state o f the art machine learning models however\, even machine learning experts do not have a clear understanding of why say a deep neural network makes a particular prediction.\nWe propose to explain the predictions of a spe cific class of ML models\, namely deep neural networks\, by pointing to t he set of what we call representer points in the training set\, for a giv en test point prediction.\nSpecifically\, we show that we can decompose t he preactivation prediction of a neural network into a linear combination of activations of training points\, with the weights corresponding to wh at we call representer values\, which thus capture the importance of that training point on the learned parameters of the network.\nOur method is scalable enough to allow for real-time explanations and feedback.\n　(Jo int work with Chih-Kuan Yeh\, Joon Sik Kim\, Ian En-Hsu Yen)\n\n関連リ ンク：\nAssociate Professor Pradeep Ravikumar's Home Page \n\n\n\n\n16 :45 - 17:00\n\n総括、閉会挨拶：\n産業技術総合研究所 人 工知能研究センター　麻生 英樹 副研究センター長 \n\n \n\n\n\n\n LOCATION:テレコムセンタービル東棟14階 MONO研修室 〒135-006 4 東京都江東区青海2-5-10 URL:https://techplay.jp/event/701429?utm_medium=referral&utm_source=ics&utm _campaign=ics END:VEVENT END:VCALENDAR