Hisashi Kashima 研究室

主宰者：Hisashi Kashima

京都大学

AI 要約（直近 5 年の研究成果）

Kashima研究室では、機械学習と統計的推論の基盤的な理論・手法を開発し、教育、医療、交通管理などの実社会問題に適用する研究を展開しています。特に、不完全または限定的な情報から正確な予測や判断を行うことが中心的なテーマです。例えば、学生の成績評価では試験問題に誤りが見つかった後の再採点方法、大規模なデータセットにおける因果推論では隠れた交絡要因を考慮した個別の治療効果推定、そして教育データの効率的な収集では労働者の信頼性が異なるクラウドソーシングからの学習に取り組んでいます。データ駆動社会における重要な課題として、モデル評価の信頼性向上にも注力しています。大規模言語モデルを用いた自動採点で人間の評価との乖離を埋める枠組み、認知バイアスの検出と緩和、そして複数の基準を考慮した多角的な評価方法を開発しています。また時系列データの解析では、分類に有効な部分パターンの自動学習や、時系列変化の説明可能な予測法を提案しており、交通流予測など実務的な応用にも活かされています。全体として、確率的な推論、因果推論、転移学習、グラフニューラルネットワークなど多様な機械学習手法を組み合わせながら、ノイズや不確実性を含む現実的なデータから、信頼性の高い意思決定を支援する基盤技術を構築することが研究室の主要な目標です。

※ AI（Claude）が、公開されている論文要旨から研究の問い・手法・主要な発見を事実情報として抽出・再構成して自動生成しています。誤りを含む可能性があるため、正確性は研究室公式情報でご確認ください。

外部リンク

研究成果（53 件）

[2026] Robust Post-hoc Score Allocation in Exams
DOI: https://doi.org/10.1145/3785022.3785027
[2026] Dynamic Feature Selection on Variable Feature Sets using Features of Features
DOI: https://doi.org/10.1109/icassp55912.2026.11463442
[2026] MAPLE: Multi-Aspect Panels of LLM Evaluators for Open-Ended Questions
DOI: https://doi.org/10.18653/v1/2026.findings-acl.1351
[2026] Estimating Learners' Skill Acquisition Without Temporal Information
[2025] Learning Location-Guided Time-Series Shapelets
DOI: https://doi.org/10.1109/tkde.2025.3536462
[2025] Scalable individual treatment effect estimator for large graphs
DOI: https://doi.org/10.1007/s10994-024-06694-w
[2025] Cognitive Biases in Large Language Models: A Survey and Mitigation Experiments
DOI: https://doi.org/10.1145/3672608.3707812
[2025] Flexible Source-free Domain Generalization via Domain Prompt-Discriminator Collaborative Learning
DOI: https://doi.org/10.1109/ijcnn64981.2025.11228941
[2025] Federated Source-Free Domain Adaptation for Classification: Weighted Cluster Aggregation for Unlabeled Data
DOI: https://doi.org/10.1109/wacv61041.2025.00669
[2025] Transfer learning with pre-trained conditional generative models
DOI: https://doi.org/10.1007/s10994-025-06748-7

続きを表示（残り 43 件）

[2024] Learning Counterfactual Explanations with Intervals for Time-series Classification
DOI: https://doi.org/10.1145/3627673.3679952
[2024] Label Selection Approach to Learning from Crowds
DOI: https://doi.org/10.1527/tjsai.39-5_f-o23
[2024] Meta graphical lasso: uncovering hidden interactions among latent mechanisms
DOI: https://doi.org/10.1038/s41598-024-68959-7
[2024] Understanding and Improving Source-Free Domain Adaptation from a Theoretical Perspective
DOI: https://doi.org/10.1109/cvpr52733.2024.02694
[2024] Optimizing Human Learning. 4th Workshop eliciting Adaptive Sequences for Learning and Educational RecSys (WASL 2024)
[2024] Queue Length Prediction Using Traffic-theory-based Deep Learning
DOI: https://doi.org/10.1527/tjsai.39-2_c-n92
[2024] AHP-Powered LLM Reasoning for Multi-Criteria Evaluation of Open-Ended Responses
DOI: https://doi.org/10.18653/v1/2024.findings-emnlp.101
[2023] Time-series Shapelets with Learnable Lengths
DOI: https://doi.org/10.1145/3583780.3615082
[2023] Learning state importance for preference-based reinforcement learning
DOI: https://doi.org/10.1007/s10994-022-06295-5
[2023] QTNet: Theory-based Queue Length Prediction for Urban Traffic
DOI: https://doi.org/10.1145/3580305.3599890
[2023] Causal Effect Estimation on Hierarchical Spatial Graph Data
DOI: https://doi.org/10.1145/3580305.3599269
[2023] Behavior Estimation from Multi-Source Data for Offline Reinforcement Learning
DOI: https://doi.org/10.1609/aaai.v37i9.26326
[2023] Multiview Representation Learning from Crowdsourced Triplet Comparisons
DOI: https://doi.org/10.1145/3543507.3583431
[2022] Chemical property prediction under experimental biases
DOI: https://doi.org/10.1038/s41598-022-12116-5
[2022] Making individually fair predictions with causal pathways
DOI: https://doi.org/10.1007/s10618-022-00885-6
[2022] InfoCEVAE: treatment effect estimation with hidden confounding variables matching
DOI: https://doi.org/10.1007/s10994-022-06246-0
[2022] GraphITE: Estimating Individual Effects of Graph-structured Treatments
DOI: https://doi.org/10.1527/tjsai.37-2_d-m73
[2022] Twin Papers
DOI: https://doi.org/10.1145/3511808.3557716
[2022] Improving Pairwise Rank Aggregation via Querying for Rank Difference
DOI: https://doi.org/10.1109/dsaa54385.2022.10032454
[2022] In-Vehicle Network Attack Detection Across Vehicle Models
DOI: https://doi.org/10.1527/tjsai.37-5_c-m42
[2022] Mitigating Observation Biases in Crowdsourced Label Aggregation
DOI: https://doi.org/10.1109/icpr56361.2022.9956439
[2022] Learning Optimal Priors for Task-Invariant Representations in Variational Autoencoders
DOI: https://doi.org/10.1145/3534678.3539291
[2022] Interpretable Knowledge Tracing: Simple and Efficient Student Modeling with Causal Relations
DOI: https://doi.org/10.1609/aaai.v36i11.21560
[2022] Learning Evolvable Time-series Shapelets
DOI: https://doi.org/10.1109/icde53745.2022.00064
[2022] Counterfactual Propagation for Treatment Effect Estimation
DOI: https://doi.org/10.1527/tjsai.37-3_b-la3
[2022] Improving imbalanced classification using near-miss instances
DOI: https://doi.org/10.1016/j.eswa.2022.117130
[2022] Poincare: Recommending Publication Venues via Treatment Effect Estimation
DOI: https://doi.org/10.1016/j.joi.2022.101283
[2022] Context-aware spatio-temporal event prediction via convolutional Hawkes processes
DOI: https://doi.org/10.1007/s10994-022-06136-5
[2022] Constant Time Graph Neural Networks
DOI: https://doi.org/10.1145/3502733
[2022] Bayesian optimization with partially specified queries
DOI: https://doi.org/10.1007/s10994-021-06079-3
[2021] Predicting anesthetic infusion events using machine learning
DOI: https://doi.org/10.1038/s41598-021-03112-2
[2021] GraphITE
DOI: https://doi.org/10.1145/3459637.3482349
[2021] Leveraging Crowdsourcing to Detect Improper Tasks in Crowdsourcing Marketplaces
DOI: https://doi.org/10.1609/aaai.v27i2.18987
[2021] A Convex Formulation for Learning from Crowds
DOI: https://doi.org/10.1609/aaai.v26i1.8105
[2021] Multinomial Relation Prediction in Social Data: A Dimension Reduction Approach
DOI: https://doi.org/10.1609/aaai.v26i1.8110
[2021] In-Vehicle Network Attack Detection Across Vehicle Models: A Supervised-Unsupervised Hybrid Approach
DOI: https://doi.org/10.1109/itsc48978.2021.9564896
[2021] Deep Mixture Point Processes
DOI: https://doi.org/10.1527/tjsai.36-5_c-l37
[2021] Grab the Reins of Crowds: Estimating the Effects of Crowd Movement Guidance Using Causal Inference
DOI: https://doi.org/10.65109/yncx8588
[2021] Student-t Variational Autoencoder for Robust Multivariate Density Estimation
DOI: https://doi.org/10.1527/tjsai.36-3_a-ka4
[2021] Learning Individually Fair Classifier with Path-Specific Causal-Effect Constraint
[2021] An Intransitivity Model for Matchup and Pairwise Comparison
DOI: https://doi.org/10.1109/icpr48806.2021.9412462
[2021] Computationally Efficient Wasserstein Loss for Structured Labels
DOI: https://doi.org/10.18653/v1/2021.eacl-srw.1
[2021] Learning to Rank for Multi-Step Ahead Time-Series Forecasting
DOI: https://doi.org/10.1109/access.2021.3068895

科研費（0 件）

まだデータがありません（KAKEN 取り込み後に表示）。

所属学会・役職（0 件）

まだデータがありません（学会データ連携後に表示）。

AI 要約（直近 5 年の研究成果）

外部リンク

関連研究室(8 件)

研究成果（53 件）

科研費（0 件）

所属学会・役職（0 件）