SIGNATE Co., Ltd.
Let’s try air battle AI using a simulator! Recruiting participants for “3rd Air Combat AI Challenge”
~The Defense Equipment Agency’s machine learning contest, which attracted much attention last time, will start on December 1st this year~
……
SIGNATE Co., Ltd. (Headquarters: Chiyoda-ku, Tokyo, President: Hide Saito, pronunciation: SIGNATE, hereafter SIGNATE), which provides total support for corporate DX from development to training, is a company that provides defense equipment with its data science platform “SIGNATE”. The “3rd Air Combat AI Challenge – Let’s try air combat AI using a simulator!” sponsored by the Agency will start on Friday, December 1st of this year.
This competition is the third edition of the “Air Combat AI Challenge” machine learning contest sponsored by the Defense Equipment Agency, and is a perfect opportunity to test your deep reinforcement learning skills while participating in the Defense Equipment Agency’s AI research. It has become an opportunity. Please also post your enthusiasm for participating and the progress of your posts on X (formerly Twitter). The hashtag is “#Air Combat AI Challenge”. We look forward to your participation.
[Image 1
(https://signate.jp/competitions/769/leaderboard) ■Competition Overview – Event Overview – Name: 3rd Air Battle AI Challenge ~ Let’s try air battle AI using a simulator! ～Sponsor: Defense Equipment Agency Prize money: 1st place ¥700,000, 2nd place ¥500,000, 3rd place ¥300,000, 4th place ¥150,000, 5th place ¥50,000, Special award ¥100,000 × up to 3 people Eligibility: Competition participation rules and special terms Submission period: 2023/12/1 (Friday) – 2024/2/25 (Sunday) Site: https://bit.ly/3SOEAb6-Assignment-Participants in this competition will receive a 2-aircraft formation ( Participants will be asked to submit behavioral judgment models that judge the behavior of a single model) and compete to see which model is superior. -Points that made it easier to participate-
Changes from last time include changing from a battle between four aircraft formations (two types of aircraft) to a battle between two aircraft formations (single model), and changing the setting to shorter battles at slightly closer range. This makes it easier to do trial and error in a shorter amount of time than before, making it a simple and easy-to-approach subject. Continuing from last time, we are also offering a free online course where you can learn the basics of reinforcement learning, creating an environment where even
first-timers can take on the challenge. – Information Disclosure Policy – Code, trained models, and data analysis results can be disclosed on the SIGNATE forum where all participants can see them. Furthermore, after the end of this competition and only for
non-commercial purposes, participants will be able to publish their models and analysis results within the specified content and scope. Please check the competition page for details. ■What is SIGNATE (https://signate.co.jp/) SIGNATE is a venture company that aims to support the promotion of DX in society and companies and contribute to the growth of Japan. Japan’s largest data talent community “SIGNATE” (https://signate.jp/company_about) provides opportunities for individuals to demonstrate their talents through AI development competitions and DX project support, as well as a human resources development cloud for promoting DX. The service “SIGNATE Cloud” (https://cloud.signate.jp/) provides education to turn employees into DX personnel. ■Company Profile Company name: SIGNATE Inc. Head office: KANDA SQUARE (within WeWork), 2-2-1 Kanda Nishikicho, Chiyoda-ku, Tokyo Capital: 280 million yen (including capital reserve of 180 million yen/2022) As of August 1, 2018) Representative: Hide Saito, Representative Director and President URL:
https://signate.co.jp/■Inquiry regarding Press Release SIGNATE Co., Ltd. Public Relations Officer Tanabe Tel: 03-4361-2454 E-mail: info@signate.co.jp