Hiroki Morise, Kyohei Atarashi, Satoshi Oyama and Masahito Kurihara: "Neural collaborative filtering with multicriteria evaluation data", Journal of Applied Soft Computing, Vol 119, 2022.
Kyohei Atarashi, Satoshi Oyama, and Masahito Kurihara: "Factorization Machines with Regularization for Sparse Feature Interactions", Journal of Machine Learning Research, Vol 22, No 1, pp. 6783--6832, 2021.
# 特徴の交互作用を扱うモデルであるFactorization Machinesにおいて，予測に重要な組合せを選択する「組合せ特徴選択」を行う方法を提案しました。
Implementation: https://github.com/neonnnnn/sparsepoly
Kyohei Atarashi, Satoshi Oyama, and Masahito Kurihara: "Sparse Random Feature Maps for the Item-multiset Kernel", Neural Networks, Vol 143, pp. 500-514, 2021.
# 特徴の交互作用を扱うカーネル関数であるItemsetカーネルを拡張した、Item-multisetカーネルに対する、高速かつ省メモリなランダム特徴を提案しました。
Kyohei Atarashi, Satoshi Oyama, and Masahito Kurihara: "Link Prediction Using Higher-order Feature Combinations across Objects", IEICE Transactions on Information and Systems, E103-D, No 8, pp. 1833-1842, August 2020.
# 特徴ベースのリンク予測において，二例間の高次の特徴の組合せを扱うモデルと効率的な評価・最適化アルゴリズムを提案しました
Kyohei Atarashi and Masakazu Ishihata: "Vertical federated learning for higher-order factorization machines". In Proceedings of the 25th Pacific-Asia Conference on Knowledge Discovery and Data Mining (PAKDD2021), pp. 346--357, 2021.
Kazuki Toda, Kyohei Atarashi, Satoshi Oyama, and Masahito Kurihara: "Unsupervised Feature Learning for Output Control of Generative Models", In Proceedings of Joint 11th International Conference on Soft Computing and Intelligent Systems and 21st International Symposium on Advanced Intelligent Systems (SCIS-ISIS), 2020.
Kazuma Itakura, Kyohei Atarashi, Satoshi Oyama, and Masahito Kurihara: "Adapting the Learning Rate of the Learning Rate in Hypergradient Descent ", In Proceedings of Joint 11th International Conference on Soft Computing and Intelligent Systems and 21st International Symposium on Advanced Intelligent Systems (SCIS-ISIS), 2020.
Kyohei Atarashi, Subhransu Maji and Satoshi Oyama:"Random Feature Maps for the Itemset Kernel", In Proceedings of the 33rd AAAI Conference on Artificial Intelligence (AAAI-19), pp. 3199--3206, January 2019. (Acceptance rate = 1150/7095 = 16.2%)
Erratum: Z_{RK}(x) = \frac{1}{m\sqrt{D}} \sum_{t=1}^m a^{m-t} \circ (\Omega^{\circ t}x^{\circ t}) -> Z_{RK}(x) = \frac{1}{m\sqrt{D}} \sum_{t=1}^m (-1)^{t+1} a^{m-t} \circ (\Omega^{\circ t}x^{\circ t}) in Equation (13) (I forgot to multiply (-1)^{t+1}).
# 特徴の交互作用を扱うカーネル関数に対してランダム特徴を提案しました
Implementation: https://github.com/neonnnnn/pyrfm
*Kyohei Atarashi, *Akimi Moriyama, Satoshi Oyama and Masahito Kurihara: "A Personalized Affect Response Model for Online News Articles", In Proceedings of the Fifth Linguistic and Cognitive Approaches To Dialog Agents Workshop (LaCATODA 2019), in conjunction with the 28th International Joint Conference on Artificial Intelligence (IJCAI 2019), August 2019.
*: alphabetical order
Kyohei Atarashi, Satoshi Oyama, and Masahito Kurihara: "Semi-supervised Learning from Crowds Using Deep Generative Models", In Proceedings of the 32nd AAAI Conference on Artificial Intelligence (AAAI-18), pp. 1555--1562, February 2018. (Acceptance rate = 933/3800 = 25%)
# 大量の「ラベルなしデータ」と少数の「クラウドソーシングのワーカによって与えられた誤りを含むラベル付きデータ」から予測モデルを構築する方法を提案しました
Yukino Baba, Tomoumi Takase, Kyohei Atarashi, Satoshi Oyama, and Hisashi Kashima: "Data Analysis Competition Platform for Educational Purposes: Lessons Learned and Future Challenges", In Proceedings of the Eighth AAAI Symposium on Educational Advances in Artificial Intelligence (EAAI-18), pp. 7887--7892, February 2018.
Kyohei Atarashi, Satoshi Oyama, Masahito Kurihara and Kazune Furudo: "A Deep Neural Network for Pairwise Classification: Enabling Feature Conjunctions and Ensuring Symmetry", In Proceedings of the 21st Pacific-Asia Conference on Knowledge Discovery and Data Mining (PAKDD2017), Lecture Notes in Computer Science, Vol. 10234, pp. 83--95, Springer, May 2017. (Acceptance rate = 129/459 = 28.2%)
# 二つのオブジェクトが同一か否かを判定する問題であるペアワイズ分類のために，二例間の特徴の組合せを扱う深層学習ベースの方法を提案しました
馬場 雪乃, 高瀬 朝海, 新 恭兵, 小山 聡, 鹿島 久嗣： "教育用データ解析コンペティション基盤の設計と実践", 情報処理学会デジタルプラクティス, Vol.9, No. 4, pp.859-873, 2018 (Invited paper).