1. Home
  2. Research
  3. Centers & Labs
  4. RIKEN Center for Advanced Intelligence Project
  5. Artificial Intelligence in Society Research Group

RIKEN Center for Advanced Intelligence Project Business and Economic Information Fusion Analysis Team

Team Leader: Takahiro Hoshino (Ph.D.)

Research Summary

Takahiro  Hoshino(Ph.D.)

In the era of rapid technological innovation and high social and economic uncertainty, government and companies are required to make decisions more quickly than ever. Although various large big-data such as transaction logs and location information, various studies showed that they are not useful for managerial or policy decision making as it is, because the big-data suffer from various biases such as selection bias. This team will develop new data-fusion techniques for various types of datasets including governmental survey data, big-data and macro-level information, to improve accuracy of public statistical information, or to aid investment/managerial decision making. We also investigate new data acquisition methods in business and economic fields which utilize statistical machine learning methods.

Research Subjects:

  • Development of Data fusion techniques
  • Development of new data acquisition methods in business and economic fields
  • Inference with anonymization of big-data in business and economics fields

Main Research Fields

  • Economics & Business

Related Research Fields

  • Neuroscience & Behavior
  • Computer Science
  • Mathematics
  • Social Sciences & General

Selected Publications

Papers with an asterisk(*) are based on research conducted outside of RIKEN.

  • 1. Shinoda, K., and Hoshino, T.:
    "Estimation of Local Average Treatment Effect by Data Combination".
    AAAI2022, Proceedings of the AAAI Conference on Artificial Intelligence (2022)
  • 2. Miyazaki, K., Hoshino, T., and Bockenholt, U:
    "Dynamic two stage modeling for category-level and brand-level purchases using potential outcome approach with semiparametric Bayes inference".
    Journal of Business & Economic Statistics, 39(3), 622-635(2021)
  • 3. Kato, T., and Hoshino, T.:
    "Semiparametric Bayesian multiple imputation for regression models with missing mixed continuous-discrete covariate".
    Annals of the Institute of Statistical Mathematics, 72(3), 803-825. (2020).
  • 4. Mitsuhiro, M., and Hoshino, T.:
    "Kernel Canonical Correlation Analysis for Data Fusion of Multiple Source Datasets".
    Japanese Journal of Statistics and Data Science, 3(2), 669-691. (2020)
  • 5. Takahata, K., and Hoshino, T.:
    "Parametric Identification of the Joint Distribution of the Potential Outcomes".
    Stat, 9(1), e254. DOI:10.1002/sta4.254
  • 6. Kato, R., and Hoshino, T.:
    "The impact of competitors store flyer advertisement on EDLP/HiLo chain performance in highly competitive retail market: GPS information and POS data approach in Japan".
    Journal of Advertising, 48(5), 569-587.(2020)
  • 7. Shimizu, Y., and Hoshino, T.:
    "Doubly Robust-type Estimation of Population Moments or Parameters in Biased Sampling".
    Stat, 8(1), e241. (2020)
  • 8. Igari, R., and Hoshino, T.:
    "Bayesian Data Combination Approach for Repeated Durations under Unobserved Missing Indicators"
    Computational Statistics & Data Analysis, 126, 150-166. (2018)
  • 9. *Okada,K. and Hoshino, T.:
    "Researchers’Choice of Number and Range of Levels in Experiments Affects the Resultant Variance-Accounted-For Effect Size".
    Psychonomic Bulletin & Review, 24(2), 607-616 (2017)
  • 10. *Hoshino, T.:
    "Semiparametric Bayesian Estimation for Marginal Parametric Potential Outcome Modeling: Application to Causal Inference".
    Journal of the American Statistical Association, 108, 1189-1204.(2013)

Related Links

Lab Members

Principal investigator

Takahiro Hoshino
Team Leader

Core members

Junichiro Niimi
Visiting Scientist
Ryosuke Igari
Visiting Scientist
Daisuke Moriwaki
Visiting Scientist
Rin Futara
Research Part-time Worker I
Kaoru Babasaki
Research Part-time Worker II
Taiga Hashimoto
Research Part-time Worker II

Contact Information

2-15-45, Mita,
Minato-ku, Tokyo, 108-8345
Email: bayesian [at] jasmine.ocn.ne.jp