本文Reparameterized Sampling for Generative Adversarial Networks被ECML-PKDD 2021接收,并获得Best (Student) Machine Learning Paper奖。 会议简介:The European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases(ECML-PKDD)是欧洲顶级的机器学习与数据挖掘会议。因为疫情原因...
蚂蚁AI Infra 团队在深度学习最核心之一的优化器方向持续投入与创新,实现了 AI 训练节约资源、加速收敛、提升泛化等目标。我们将推出“优化器三部曲”系列,这是本系列的第二篇。 在搜索、推荐、广告场景中,充…
MACHINE learningMachine Learning -doi:10.1007/s10994-021-06062-yAppice, AnnalisaDipartimento di Informatica, Università degli Studi di Bari “Aldo Moro”, Bari, ItalyEscalera, SergioUniversitat de Barcelona and Computer Vision Center, Barcelona, SpainGámez, Jose A....
The European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases will take place in Bilbao, Basque Country (Spain), from the 13th to the 17th of September 2021. This event is the premier European machine learning and data mining conference and builds upon...
https://2021.ecmlpkdd.org/ *SCOPE AND MOTIVATION* This event is the premier European machine learning and data mining conference. The papers accepted for the journal track will be published in the Machine Learning Journal and the Data Mining and Knowledge Discovery Journal. Papers on all topics...
Introduction to the special issue of the ECML PKDD 2021 journal trackDATA miningTIME series analysisMACHINE learningData Mining and Knowledge Discovery -doi:10.1007/S10994-021-06062-YAnnalisa AppiceSergio EscaleraJose A. GámezHeike TrautmannSpringer USData Mining and Knowledge Discovery...
ECML PKDD 2021|PaGNN:基于交互结构学习的链路预测 本文将解读蚂蚁集团图学习团队在ECML PKDD 2021上发表的论文《Inductive Link Prediction with Interactive Structure Learning on Attributed Graph》。该论文提出的路径感知图神经网络PaGNN主要探索了在链接预测任务中如何将节点间的复杂网络结构学习集成到传统GNN的计算...
Adaptive Supervised Learning for Volatility Targeting Models (Ecml Pkdd Midas 2021 Presentation Slides)doi:10.2139/ssrn.3926218Volatility targetingSupervised learningBest orderingModel-based and Portfolio allocationWalk-forward and Features selection.In the context of risk-based portfolio construction and pro-...
本文将解读蚂蚁集团图学习团队在ECML PKDD 2021上发表的论文《Inductive Link Prediction with Interactive Structure Learning on Attributed Graph》。该论文提出的路径感知图神经网络PaGNN主要探索了在链接预测任务中如何将节点间的复杂网络结构学习集成到传统GNN的计算范式中,从而提高链接预测的准确率。目前,PaGNN已成功应...
Multi-View Self-Supervised Heterogeneous Graph Embedding, 2021, ECML-PKDD 本文提出一种新的基于元路径的异构图自监督学习方法MVSE,采用两个视图intra-view和inter-view的对比学习。 实验的数据集是DBLP、ACM、IMDB。 2 方法论 2.1 模型总览 给定异构图中的一个结点,MVSE首先采样其基于元路径的子图,然后通过semant...