A team which included two data scientists from Nikkei achieved gold medals in the machine-learning competition "HMS - Harmful Brain Activity Classification" hosted by Kaggle, the world's largest data science community.

The 5-member team ranked 14th out of 2767 entries, and included Kazuaki Tokunaga and Takuto Oshima from Nikkei's Data Intelligence Group, Platform Business Office. Tokunaga and Oshima were also awarded the title of "Kaggle Master."

The competition's theme was the classification of brain signal patterns, and it was held from 10th January to 9th April 2024. (https://www.kaggle.com/competitions/hms-harmful-brain-activity-classification) The goal was to accurately identify signs of harmful brain activity such as epilepsy from electroencephalography signals and spectrogram images recorded from patients.

Effective use of image processing models was key to improving prediction accuracy. Tokunaga and Ohshima's work involves analyzing data from articles and advertisements, including images, in order to improve products such as the Nikkei online edition. Winning this award is testament to Nikkei's high level of image processing technology. The knowledge and experience gained from the competition will be applied in their work going forward.

This is the fourth time for Nikkei employees to be honored with gold medals in a Kaggle competition, and Nikkei currently has six Kaggle Masters as employees.

(About Kaggle)
Kaggle is the world's largest online community of users focusing on data science and machine learning, and its members come from around the world. Competitions held by Kaggle are popular events that target the predictive performance of machine learning models. The title of "Kaggle Master" is considered to represent the top 1% of active users.

https://www.nikkei.co.jp/nikkeiinfo/en/news/announcements/1034.html