arXivDaily arXiv每日学术速递 周一至周五更新

期刊&会议

Transactions on Machine Learning Research · 期刊 · Machine Learning

2025-12-03 至 2025-12-03 共收录 5
2511.21890 2025-12-03 stat.ML cs.LG

Sparse Multiple Kernel Learning: Alternating Best Response and Semidefinite Relaxations

稀疏多核学习:交替最佳响应与半正定松弛

Dimitris Bertsimas, Caio de Prospero Iglesias, Nicholas A. G. Johnson

机构 * Massachusetts Institute of Technology(麻省理工学院)

AI总结 本文提出了一种稀疏多核学习方法,通过交替最佳响应算法和半正定松弛,实现高效且精确的核选择与优化。

Comments Transactions on Machine Learning Research (2025)

详情

展开后加载摘要…

URL PDF HTML 收藏
2504.19785 2025-12-03 cs.LG

Heterophily-informed Message Passing

异质性引导的消息传递

Haishan Wang, Arno Solin, Vikas Garg

AI总结 本研究提出一种基于异质性的消息传递方法,通过调节消息聚合以保留信息的低频和高频成分,提升图神经网络在分类和分子生成任务中的性能。

Comments Appearing in Transactions on Machine Learning Research (TMLR) 2025

Journal ref Transactions on Machine Learning Research (2025)

详情

展开后加载摘要…

URL PDF HTML 收藏
2408.01402 2025-12-03 cs.LG cs.AI cs.CL

Pre-trained Language Models Improve the Few-shot Prompt Ability of Decision Transformer

预训练语言模型提升决策变换器的少样本提示能力

Yu Yang, Pan Xu

机构 * Department of Electrical and Computer Engineering(电气与计算机工程系) Duke University(杜克大学) Department of Biostatistics and Bioinformatics(生物统计学与生物信息学系) Department of Computer Science(计算机科学系)

AI总结 预训练语言模型提升决策变换器的少样本提示能力,通过初始化和提示正则化增强RL任务区分能力。

Comments 2 figures, 10 tables. Published in Transactions on Machine Learning Research (TMLR)

Journal ref Transactions on Machine Learning Research, 2025

详情

展开后加载摘要…

URL PDF HTML 收藏
2512.02303 2025-12-03 cs.LG q-bio.BM

Training Dynamics of Learning 3D-Rotational Equivariance

学习3D旋转等价性动力学

Max W. Shen, Ewa Nowara, Michael Maser, Kyunghyun Cho

机构 * Genentech Computational Sciences(基因泰克计算科学)

AI总结 本文研究了学习3D旋转等价性的动力学,发现其损失景观更平滑且更容易优化,从而在训练早期即可显著降低等价性误差。

Comments Accepted to Transactions on Machine Learning Research (TMLR)

详情

展开后加载摘要…

URL PDF HTML 收藏
2509.08366 2025-12-03 stat.ML cs.LG math.ST stat.ME stat.TH

kNNSampler: Stochastic Imputations for Recovering Missing Value Distributions

kNNSampler: 通过随机采样恢复缺失值分布的随机填补方法

Parastoo Pashmchi, Jérôme Benoit, Motonobu Kanagawa

机构 * SAP Labs France(SAP法国实验室) EURECOM

AI总结 kNNSampler通过随机采样最相似单元的观测响应来填补缺失值,能够估计缺失值的条件分布并用于多重填补。

Comments Published in Transactions on Machine Learning Research (TMLR). Reviewed on OpenReview: https://openreview.net/forum?id=4CDnIACCQG

详情

展开后加载摘要…

URL PDF HTML 收藏