JP

Events

Category
Date
Title
Presenter/Location
Details
2025/06/12 Thu
17:00〜18:30
Takashi Kunimoto (Singapore Management University)
本館1階会議室またはオンライン開催
2025/05/29 Thu
17:00〜18:30
TBA
Patrick DeJarnette (Waseda University)
本館1階会議室またはオンライン開催
2025/05/22 Thu
17:00〜18:30
TBA
Eric Weese (University of Tokyo)
本館1階会議室またはオンライン開催
2025/04/24 Thu
17:00〜18:30
TBA
Noriaki Kiguchi (Hitotsubashi University)
本館1階会議室またはオンライン開催
2025/04/17 Thu
17:00〜18:30
TBA
Yuya Wakabayashi (Osaka University)
本館1階会議室またはオンライン開催
2025/04/10 Thu
17:00〜18:30
TBA
Fabio Maccheroni (Bocconi University)
本館1階会議室またはオンライン開催
2025/03/27 Thu
17:00〜18:30
TBA
Kentaro Asai (Kyoto University)
本館1階会議室またはオンライン開催
2025/03/07 Fri
16:30〜18:00
Culture, tastes, and market integration: Testing the localized tastes hypothesis (with T. Mori)
Jens Wrona(University of Duisburg-Essen)
京都大学経済研究所本館1階 106 会議室
2025/02/19 Wed
16:45〜18:15
Constrained Classification and Policy Learning
坂口 翔政(東京大学)
第一共同研究室(4F 北側)
2025/02/05 Wed
16:45〜18:15
Unsupervised Learning for High-dimensional Distributions with Tree-based Methods
粟屋 直(早稲田大学)
第一共同研究室(4F 北側)

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).

TOP