-

Mitsubishi Electric Data Scientist Team Wins Gold in Kaggle Competition

Two members awarded Kaggle Competition Master titles for cumulative achievements

TOKYO--(BUSINESS WIRE)--Mitsubishi Electric Corporation (TOKYO: 6503) announced today that a team of three data scientists, Yuya Shintani, Kento Okumura and Shoma Fukuhara from the Mitsubishi Electric DX Innovation Center, won a gold medal in the Home Credit – Credit Risk Model Stability competition hosted by Google LLC subsidiary Kaggle INC., the world’s largest AI competition platform.

In addition, Messrs. Shintani and Okumura were each awarded the title of Kaggle Competition Master for cumulatively winning one gold medal and two silver medals each in Kaggle competitions. The Master title, which is presented to people who have received a certain number of awards in the competition category, has been presented to 2,108 people worldwide, including 278 in Japan (as of June 18).

Kaggle medals recognize winners for using machine learning in a competitive format to develop effective solutions to challenges posed by companies and government agencies.

The recent competition, which ran from February 5 to May 27, posed the challenge of predicting the likelihood of loan applicants defaulting on their debts, based on which participants used analytical techniques to propose solutions. The solution devised by Messrs. Shintani, Okumura and Fukuhara, which ranked 13th out of 3,856 solutions, won a gold medal for its robust features, model construction and adaptation to evaluation metrics based on exploratory data analysis, a method of data science that uses basic aggregation and visualization to explore data characteristics and understand data structure.

For the full text, please visit: www.MitsubishiElectric.com/news/

Contacts

Customer Inquiries
DX Innovation Center
Mitsubishi Electric Corporation
dx-ic@me.MitsubishiElectric.co.jp
www.MitsubishiElectric.com/

Media Inquiries
Takeyoshi Komatsu
Public Relations Division
Mitsubishi Electric Corporation
Tel: +81-3-3218-2332
prd.gnews@nk.MitsubishiElectric.co.jp
www.MitsubishiElectric.com/news/

Mitsubishi Electric Corporation

TOKYO:6503

Release Versions

Contacts

Customer Inquiries
DX Innovation Center
Mitsubishi Electric Corporation
dx-ic@me.MitsubishiElectric.co.jp
www.MitsubishiElectric.com/

Media Inquiries
Takeyoshi Komatsu
Public Relations Division
Mitsubishi Electric Corporation
Tel: +81-3-3218-2332
prd.gnews@nk.MitsubishiElectric.co.jp
www.MitsubishiElectric.com/news/

More News From Mitsubishi Electric Corporation

Mitsubishi Electric Announces Director Candidates

TOKYO--(BUSINESS WIRE)--Mitsubishi Electric Corporation (TOKYO: 6503) announced today that the candidates for Directors, who will be proposed for election at the Ordinary General Meeting of Shareholders scheduled in June 2026, were elected at the company’s Nomination Committee meeting held earlier today. Director Candidates (current positions) Tatsuro Kosaka (Director (Senior Advisor, CHUGAI PHARMACEUTICAL CO., LTD.)) Hiroyuki Yanagi (Director) Masako Egawa (Director (Chancellor, Seikei Gakuen)...

Mitsubishi Electric Invests in AI Startup Sakana AI

TOKYO--(BUSINESS WIRE)--Mitsubishi Electric Corporation (TOKYO: 6503) announced today that it has invested in Sakana AI Inc., a Japan-based AI startup that develops next‑generation AI-foundation models. Through this collaboration, Mitsubishi Electric aims to create new solutions and expand businesses related to its SerendieTM digital platform by leveraging Sakana AI’s expertise in using AI to optimize complex, tacit‑knowledge‑rich, highly challenging business operations. Due to rapid advances i...

Mitsubishi Electric Develops Edge Digital Twin Technology for Real-time Compensation of Errors in CNC Machine Tools

TOKYO--(BUSINESS WIRE)--Mitsubishi Electric Corporation (TOKYO: 6503) announced today that in collaboration with RWTH Aachen University in Germany, it has developed a technology that uses a digital twin to correct machine-tool errors in real time on machine tools equipped with computer numerical control (CNC1). Tests showed that the technology can reduce errors caused by slight deformation of machined parts due to the force of a cutting tool by up to 50%.2 This results in fewer mistakes during...
Back to Newsroom