TV Professional Baseball Commentary Using Machine Learning
We are working on a project to challenge Japanese Professional Baseball commentary using deep learning and data mining.
Phase 1 involved the learning of data that included in excess of 3 million pitches recorded in official professional baseball pitching data from 2004 to develop a digital sports commentator called Mr. ZUNO.
This system predicts the combination of deliveries, victory (or defeat) and ranking etc. using deep learning, and utilizes data mining to analyze pitching according to player tendencies and the state of play which human commentators have been unable to determine to date.
Looking towards the future, we will adapt this system to sports other than baseball and tackle the development of ways to enjoy sports from a variety of perspectives.
Naoki Tanaka (Dentsu Lab Tokyo)
Kaoru Sugano (Dentsu Lab Tokyo)
Ryuichiro Akamatsu (Dentsu)
Moe Goto (Dentsu Lab Tokyo), Yuki Otsu (Dentsu)
Ai Kohei (Dentsu Lab Tokyo), Kazuki Sakurai (Dentsu Tech), Shotaro Kamoto (Dentsu), Atsuko Okada (puzzle)
Yosuke Wakita (puzzle)
Kenichiro Takase (Data Stadium)
Hiroyuki Nagamatsu (Data Stadium)
Kei Kanazawa (Data Stadium)
Kyohei Okawa (Data Stadium), Sagiri Uehara (Data Stadium)
Nao Tokui (Qosmo, Inc.)
Kouki Yamada
Shoya Dozono (Qosmo, Inc.)
Ryuta Motegi (TWOTONE)
Yuri Matsumura (TWOTONE), Ken Hirose (TWOTONE)
Miyu Hosoi (Qosmo, Inc.)
Yusuke Koyanagi (Dentsu)
Toshiyuki Hashimoto (aircord)
Seiya Nakano (aircord)
Yumi Fukuyoshi (Taki corporation)
Yasuhide Kobayashi (Beard)
Yusuke Nishida (Dentsu Lab Tokyo)
Yukimasa Okada (puzzle)
Isao Ando (puzzle)
Shinya Matsuo
Ken Hirose (TWOTONE), Takuto Katayama (TWOTONE)
Isao Yamazaki (PINTON Co., Ltd.)
Yusuke Koyanagi (Dentsu)
Nobuto Osakabe
Kazuki Hoshi (The Elves and the Shoemaker)