机器之心 & ArXiv Weekly Radiostation参与：杜伟、楚航、罗若天本周的重要论文包括依图推出的预训练语言理解新模型 ConvBERT，以及关于机器学习模型安全性与隐私的探讨。目录：Graph Neural Networks in Recommender Systems: A SurveyKnowledge Distillation in Wide Neural Networks: Risk Bound, Data Efficiency and Imperfect TeacherConvBERT: Improving BERT with Span-based Dynamic ConvolutionA Survey of Label-noise Representation Learning: Past, Present and FutureStructured Models for Vision-and-Language ReasoningSecurity and Privacy of Machine Learning Models: A SurveryDeep Learning for Detecting Cerebral Aneurysms with CT AngiographyArXiv Weekly Radiostation：NLP、CV、ML 更多精选论文（附音频）论文 1：Graph Neural Networks in Recommender Systems: A Survey作者：Shiwen Wu、Wentao Zhang、Fei Sun、Bin Cui论文链接
………………………………
