Thời gian: 14h~18h Thứ Bảy - 13/04/2019 (mở cửa đón khách 13h40 ~ 13h55)
👉 Form đăng ký tham dự: https://forms.gle/rXFu9H4i8aga3Mxi6 (Hạn cuối đăng ký là 23:59 ngày 8/4/2019)
Tiếp nối thành công của VJAI Minicourse, VJAI meetup sẽ trở lại vào ngày 13/4/2019. Nội dung của buổi Meetup #15 như sau:
14:00 – 15:00 Deep learning in healthcare: Opportunities and challenges with electronic health records (EHR) data. (Trần Quang Thiện, University of Tsukuba) 15:00 – 16:00 Machine learning aided biomolecular design: from understanding functions to virtual screening (Dr. Trần Phước Duy, Tokyo Institute of Technology) 16:00 – 16:15 Break 16:15 – 16:30 Introduction to AI research at MILA (Nguyen Xuân Phong, Hitachi & Tokyo Techies) 16:30 – 18:00 Panel Discussion with Special Guest Dr. Bùi Hải Hưng (VinAI & Deepmind)
🏢 Địa điểm: Rakuten Inc. 〒158-0094 東京都世田谷区玉川１丁目１４−１ ℹ️ Access: 東急田園都市線、東急大井町線「二子玉川」駅より徒歩4分
Thông tin về địa chỉ và cách đi: https://corp.rakuten.co.jp/about/map/crimsonhouse/
Ngoài ấn Going trên event thì các bạn nhớ điền vào form dưới đây cho đến trước 23:59 ngày 8/4/2019. https://forms.gle/rXFu9H4i8aga3Mxi6
Nếu số người điền form đăng ký nhiều hơn sức chứa của hội trường, ban tổ chức sẽ áp dụng hình thức Lottery như đã thông báo trên Facebook group trong link dưới đây: https://www.facebook.com/groups/1332064783547219/permalink/1961572643929760/
Abstract của hai bài nói của Dr. Trần Phước Duy và Trần Quang Thiện:
=============================
Dr. Tran Phuoc Duy School of Life Sciences and Technology, Tokyo Institute of Technology Title: Machine learning aided biomolecular design: from understanding functions to virtual screening Abstract: The time scale of biological process, which involves upper thousands of molecules, are usually larger than microsecond exceeding the capability of the current computational approach to yield a reliable insight massively. Machine learning plays the roll as a powerful tool to help the current computational methods overcome not only the time-scale obstacle, but also help to determine the design problem for a given function in need. Here we will present and discuss about three main research themes that we are focused on: taking advantage of Markov state modeling for functional understanding of biomolecular complexes; Monte Carlo tree search as a powerful tool for enhanced sampling; actor-critic functional design of antimicrobial and antifreeze peptides.
=============================
Trần Quang Thiện Department of Computer Science, University of Tsukuba (Master’s student) Title: Deep learning in healthcare: Opportunities and challenges with electronic health records (EHR) data. Abstract: Interested in deep learning for healthcare has grown strongly recent years besides with the successes in other domains such as Computer Vision, Natural Language Processing, Speech Recognition and so forth. This talk will try to give a brief look into the recent effort of research in deep learning for healthcare. Especially, this talk focuses on the opportunities and challenges in using electronic health records (EHR) data, which is one of the most important data sources in healthcare domain.