Shinya SUZUMURA.

Affilication:
LY Corporation (since 2023). Yahoo Japan Corporation (2016 - 2023).
I am an engineer for Yahoo! JAPAN Ads (Display Ads) in LY corporation.
The main components I have developed are: (i) machine learning models for CTR and CVR predictions. (ii) auto-optimization of ad designs.

Education:
Ph.D. at Nagoya Institute of Technology (from april 2013 to march 2021 where the leave of absence is for two years). BA & MA at Cubu University (2007 - 2013).
E-mails: ssuzumur[at]lycorp.co.jp, suzumura.mllab.nit[at]gmail.com

Conference papers

  1. Shogo Iwazaki and Shinya Suzumura, "No-Regret Bandit Exploration based on Soft Tree Ensemble Model,"
    NeurIPS 2024 (research track, acceptance rate 25.8%), PDF.
  2. Ryohei Emori, Shinya Suzumura, Nobuyuki Shimizu and Takahiro Hoshino, "Leveraging Instrumental Variables in Online Advertising Auctions : Robust Click-Through-Rate Prediction,"
    AdKDD 2024, URL.
  3. Shinya Suzumura, Kazuya Nakagawa, Yuta Umezu, Koji Tsuda, Ichiro Takeuchi, "Selective Inference for Sparse High-Order Interaction Models,"
    ICML 2017 (oral presentation, acceptance rate 25.4%), PDF, supp.
  4. Kazuya Nakagawa, Shinya Suzumura, Mahito Sugiyama, Koji Tsuda, Ichiro Takeuchi, "Safe Pattern Pruning: An Efficient Approach for Predictive Pattern Mining,"
    ACM SIGKDD 2016 (research track, acceptance rate 18%), PDF.
  5. Shinya Suzumura, Kohei Ogawa, Masashi Sugiyama, Ichiro Takeuchi, "Outlier Path: A Homotopy Algorithm for Robust SVM,"
    ICML 2014 (oral presentation, acceptance rate 25%), PDF.
  6. Ryohei Nakano and Shinya Suzumura, "Grammatical Induction with Error Correction for Deterministic Context-free L-systems,"
    WCECS 2012, PDF.
  7. Shinya Suzumura and Ryohei Nakano, "Complex-valued Multilayer Perceptron Search Utilizing Eigen Vector Descent and Reducibility Mapping,"
    ICANN 2012, Splinger Link.

Journals

  1. Shinya Suzumura, Kazuya Nakagawa, Yuta Umezu, Koji Tsuda, Ichiro Takeuchi, "Selective Inference for High-order Interaction Features Selected in a Stepwise Manner,"
    IPSJ Transactions on Bioinformatics, 2021, PDF.
  2. Shinya Suzumura, Kohei Ogawa, Masashi Sugiyama, Masayuki Karasuyama, Ichiro Takeuchi, "Homotopy Continuation Approaches for Robust SV Classification and Regression,"
    Machine Learning, 2017, PDF.
  3. Shinya Suzumura and Ryohei Nakano, "Complex-Valued BFGS Method for Complex-Valued Neural Networks,"
    EICE Transactions on Information & Systems, 2013, PDF.