＜Workshop on Search and Platform ご案内＞

日時：2025年3月21日（金）9:55 ～ 17:15

会場：京都大学経済研究所北館N101、N102（京都市左京区吉田本町 京都大学吉田キャンパス本部構内）
　　　経済研究所へのアクセスはこちらを参照ください。

プログラムのURL：こちらをご参照ください。

参加費：無料

定員：50人
言語：英語 ≪同時通訳はありません≫

オーガナイザー：
渡辺誠： CIGS 上席研究員

＜登録方法＞
会議に参加希望の方は、以下の申込フォームからご登録ください。（締切：3月13日）

登録が完了すると、自動返信メールが届きます。

https://forms.gle/Eni9cjqh68pK7sfUA

メールが届かない場合は、迷惑メールフォルダをご確認ください。

登録後に内容を変更したい場合は、詳細な情報をメールでお知らせください。

＊資料の配布はございませんので、必要な場合はご自身の端末をご持参いただき、
　上記プログラムのURLよりPaper /Slidesにアクセスしてください。
　Slideは随時更新をしていきます。

よろしくお願いいたします。

南幸子

一般財団法人キヤノングローバル戦略研究所
〒100-6511　東京都千代田区丸の内1-5-1　
　　　　　　　　 新丸の内ビルディング11F
TEL:03-6213-0550　FAX:03-3217-1251
minami.sachiko@canon-igs.org

——————————————————————————-

“The Workshop on Search and Platform”
Date：2025, March 21st（Friday）9:55 ～ 17:15

Location：Room N101/N102, KIER North Building, Kyoto University, Yoshidahonmachi, Sakyo Ward, Kyoto, 606-8317

                 Here is the URL to the page of access to the Institute of Economic Research.

Participation Fee: Free
Capacity: 50 participants *The registration will be closed if we receive too many applications.

Organizer:

Makoto Watanabe (CIGS senior researcher)

Here is the URL to the agenda of the conference.

<How to register>
If you wish to participate in this conference, please register through the following application form. (Application Deadline: March 13th)

https://forms.gle/Eni9cjqh68pK7sfUA

After your registration is complete, you will receive an auto-reply from the form.

If you do not receive the email, please check your spam mail folder.

If you want to change some items after your registration, please email us with the detailed information.

[CIGS Secretariat] (Reception hours: 9:00-17:30 on weekdays)
TEL: +81-3-6213-0550 (Representative)
Mail: minami.sachiko@canon-igs.org

プログラム2024年度official (1)

Abstract
Estimating distributional structures, such as density estimation and two-sample comparison, is
a fundamental task in data science. However, estimating high-dimensional distributions is widely
recognized as challenging due to the well-known curse of dimensionality. In the case of supervised
learning, where one needs to estimate an unknown function often defined on a high-dimensional
space, a common approach in statistics and machine learning is to introduce tree-based methods,
such as boosting, random forest, and Bayesian additive regression trees. These methods are known
to be effective for such challenging tasks with feasible computation costs. This presentation aims
to introduce their counterparts for unsupervised learning. We first introduce a new non-parametric
Bayesian model for learning distributions by generalizing the Polya tree process, which is originally
introduced for low-dimensional density estimation. We next propose a new way of combining
multiple tree-based learners in the manner of boosting for improved empirical performance.
This is joint work with Li Ma (Duke University).