Statistical Data Science
Laboratory

Department of Industrial and Systems Engineering, Faculty of Science and Engineering, Hosei University, Tokyo, Japan / 日本語

About the Lab

We study statistics as the science of reading data correctly. Our research expresses the structure behind data as statistical models and uses them to predict what will happen next, from the lifetime of products to the ability of examinees. The laboratory belongs to the Department of Industrial and Systems Engineering, Faculty of Science and Engineering, Hosei University, led by Takenori Sakumura.

More about the lab

Research Themes

Bayesian estimation with noninformative priors
Deriving optimal Bayesian estimators under reference and Jeffreys priors for various distributions.
Reliability, degradation and accelerated life tests
Predicting product lifetime with statistical models from accelerated tests and degradation data.
Item response theory and adaptive testing
Test designs and estimation methods that measure ability accurately with few items.
Statistical modeling of real data
Analyzing data on infectious diseases, parts demand, purchases and sports with statistical models.

Research

Recent Papers (FY2024–FY2026)

  1. Ko Abe, Takenori Sakumura (2026) “Empirical Joint Estimation for Incubation Period and Infection under the Doubly Interval-Censored with Right Truncated Observation,” Journal of Computational and Graphical Statistics. [DOI]
  2. 今泉智子，作村建紀，岡田謙介 (2026) 「密度比に基づいた Haebara 法の拡張等化手法」 大学入試研究ジャーナル, 36, 114–121.
  3. Takenori Sakumura, Takemi Yanagimoto (2025) “Optimal Bayesian estimator under the reference prior in the von Mises distribution,” Communications in Statistics-Simulation and Computation, 1–13, (Online First). [DOI]
  4. Takenori Sakumura, Takemi Yanagimoto (2025) “Bayesian Estimation under the Conjugate-Gamma Prior for the Weibull Distribution,” Proceedings of the Institute of Statistical Mathematics, 73, 253–266, (in Japanese).
  5. Takenori Sakumura, Takemi Yanagimoto (2024) “Very Small Bias Observed in Bayesian Estimators even for Small Sample Sizes,” Communications in Statistics - Simulation and Computation, 53(11), 5594–5604. [DOI]

All publications

News

News about the lab, such as awards and conference presentations, is posted in Japanese.

News (in Japanese)

Contact