Hello, this is Shimizu from Public Relations.
We are now only a few days away from the end of 2017.
At the Matsuo Lab’s year-end party held the other day, more than 80 people including lab students, faculty members, GCI students, Deep Larning.jp, and others related to the Matsuo Lab gathered for a very lively party.
Everyone was interacting with each other on topics such as their own work and research.
Therefore, we would like to introduce some of our laboratory’s activities in 2017.
[2017 Activities.
Accepted in International Conferences and Journals
International Conferences, Workshops, etc.
◼︎Yusuke Iwasawa, et al.: Privacy Issues Regarding the Application of DNNs to Activity-Recognition using Wearables and Its Countermeasures by Use of IJCAI2017
◼︎Masaru Isonuma, et al. : Extractive Summarization Using Multi-Task Learning with Document Classification. EMNLP 2017
◼︎Pablo Loyola, et al.: A Neural Architecture for Generating Natural Language Descriptions from Source Code Changes. ACL2017 (short)
◼︎Shuhei Iitsuka et al.: Inferring Win-Lose Product Network from User Behavior, WI2017
◼︎Web Analytics, ISSRE, ICSE, ICLR (workshop), CVPR (workshop), EMNLP (workshop), total 10
4 papers being submitted to G1 conferences such as ◼︎CLR, etc.
Transactions of Information Processing Society of Japan,
︎◼︎ ︎Naoki Nonaka et al: Predicting popularity of contents using vector representation learned from Wikipedia’s edit history, IEICE Transactions,
︎◼︎ Masahiro Suzuki et al, A deep generative model for bidirectional generation between different modalities, Transactions of Information Processing Society of Japan,
︎◼︎ Koichiro Tamura et al, A proposal of shareholder value estimation model using ensemble model with deep learning, Transactions of Information Processing Society of Japan, ◼︎ Proposal of shareholder value estimation model using ensemble model, Transactions of the Japanese Society for Artificial Intelligence
︎◼︎ Nonaka Naoki et al: Car sales prediction using features contributing to consumer sentiment extracted from online reviews, Transactions of Information Processing Society of Japan
︎◼︎Matsuda N& Matsuo Y, Impact of MBA on Entrepreneurial Success: Do Entrepreneurs Acquire Capacity through the Program or Does MBA Only Signal Gifted Talent and Experience? Entrepreneurship & Organization Management
︎◼︎ Nagamine Koya et al.: A Proposal of Physical Burden Evaluation Method for Manual Wheelchair Running Using Acceleration Sensor, IEICE Transactions on Information and Communication Engineers
︎◼︎ Iwasawa Yusuke et al.: User Hostile Network, Transactions of Japanese Society for Artificial Intelligence
︎◼︎Yusuke Iwasawa et al: User adaptation of action recognition model based on deep learning with semi-supervised distillation, Transactions of the Japanese Society for Artificial Intelligence
etc.
independent study group
Weekly voluntary study sessions led by Matsuo Lab members are also held with 15~20 participants each time.
︎Reading sessions 115 times
︎DL Hacks (implementation) 16 times
*The number of sessions is the cumulative total of 2.5 years since the start.
*Materials can be found on the DL Hacks twitter account here.
Joint Research Achievements
︎◼︎Japan’s First Fully Automated AI Earnings Summary,a Joint Project with the Nikkei Shimbun (January 25, 2017)
︎◼︎Daikou Securities, AI to screen trades, to collaborate with the University of Tokyo (May 15, 2017)
︎◼︎Mizuho Bank and IGPI Business Analytics & Intelligence to collaborate on “Joint Research on Sophisticated Foreign Exchange Trading with AI” (June 30, 2017)
︎◼︎Collaborative Research with Creww Corporation,“Artificial Intelligence (AI) Predicts Matchmaking between Startups and Major Companies!” (July 14, 2017)
︎ ◼︎”Wealth Navi AI Asset Management Lab” established (November 22, 2017)
etc.
external activities
︎“Deep Learning” by Goodfellow et al. published online in Japanese translation (September 25, 2017).
What a surprise, it exceeded 30,000 UUs and 2,329 haters on the first day of publication!
◼︎Matsuo chairs the board of directors of the Japan Deep Learning Association, anincorporated association established (October 25, 2017).
︎◼︎NHK E-television’s “What’s Human? Introduction to Super AI” on NHK E-television .
◼︎︎ Free online course “DL4US” is now accepting applications (December 10, 2017).
More than 700 people applied on the second day of the application period, which will close on December 30.
Shenzhen Lab Trip
We visited a company in Shenzhen, China, also known as “the world’s fastest startup training center.
We would like to thank you for your exceptional support in 2017.
We would like to express our deepest gratitude to all of you for your support that has made it possible for us to achieve these results.
In the coming Year of the Dog, we will strive to discover new technologies with a keen sense of smell and promote their industrial application,
We will continue to produce many talented people from the community of “Matsuo Lab” who will energize Japanese industry in the years to come.
I sincerely wish you all good health and continued growth.
Please have a happy New Year.