点击蓝字
关注我们
AI TIME欢迎每一位AI爱好者的加入!
哔哩哔哩直播通道
扫码关注AI TIME哔哩哔哩官方账号预约直播
7月27日 14:30-16:30
14:30-15:00
周正阳
以不变应万变:面向时空数据的不变关联学习
15:00-15:30
万润哲
Experimentation Platforms Meet Reinforcement Learning: Bayesian Sequential Decision-Making for Continuous Monitoring
15:30-16:00
秦 欣
基于多样、判别表征的低资源领域泛化方法
16:00-16:30
张一帆
零成本测试时间自适应提升模型泛化性:域风险最小化
7月28日 10:30-12:00
10:30-11:00
赵辰
Towards Fair Disentanglement Online Learning for Changing Environment
11:00-11:30
胡竣峰
Graph Neural Processes for Spatio-Temporal Extrapolation
11:30-12:00
张凯科
Disentangled Representation Learning for Discrete-time Dynamic Graph
嘉宾介绍
周正阳
中国科大软件学院特任副研究员。周正阳于2023年6月在中国科学技术大学计算机系获博士学位。主要研究领域是时空数据挖掘与城市计算,近五年来以第一作者/通信作者身份发表CCF-B类以上高水平学术论文十余篇,其中CCF-A类及ICLR、IEEE Transaction等顶级会议/期刊共9篇,获中国科学院院长奖、中国科大校优博论文提名奖、国家奖学金等荣誉。
报告题目:以不变应万变:面向时空数据的不变关联学习
报告简介:在数据爆炸与人工智能快速发展的时代,时空数据学习已成为城市数字化的一项重要技术。由于城市的不断扩张和高度动态性,当前的时空预测模型倾向于在训练集和测试集之间发生分布偏移,导致分布外泛化(Out-of-distribution, OOD)的性能大大下降。然而,现有的研究很少关注时序回归任务中的OOD问题。具有周期性的时空数据,其数据序列及时空依赖关系往往均表现出阶段级的异质性,这为不变性提取带来了挑战。在本报告中,我们认为,时空依赖关系对泛化是有意义的,因而提出了一个因果时空学习框架,CauSTG,以将不变关系迁移至OOD场景。最后,报告指出了向时空数据泛化预测的未来研究趋势。
赵辰
贝勒大学助理教授。博士毕业于美国得克萨斯大学达拉斯分校计算机专业。主要研究方向为公平性学习在数据发掘,机器学习,深度学习上的研究和应用。在包括KDD,CVPR, AAAI,WWW,ICASSP等会议与期刊上发表过多篇论文,并受邀担任KDD,NeurIPS, AAAI,ICDM,AISTATS等人工智能领域顶级国际会议程序委员和审稿人。个人主页:https://charliezhaoyinpeng.github.io/homepage/
报告题目:Towards Fair Disentanglement Online Learning for Changing Environment
秦欣
中国科学院计算技术研究所助理研究员,研究方向为迁移学习与普适计算。
报告题目:基于多样、判别表征的低资源领域泛化方法。
万润哲
现为亚马逊应用科学家,博士毕业后专注于不确定性下最优决策的研究与应用,包括应果推断,多臂老虎机,以及强化学习。在顶级AI/统计/经济会议或期刊发表十余篇论文。个人主页 https://runzhewan.com/
报告题目:Experimentation Platforms Meet Reinforcement Learning: Bayesian Sequential Decision-Making for Continuous Monitoring
报告简介:With the growing needs of online A/B testing to support the innovation in industry, the 、opportunity cost of running an experiment becomes non-negligible. Therefore, there is an increasing demand for an efficient continuous monitoring service that allows early stopping when appropriate. Classic statistical methods focus on hypothesis testing and are mostly developed for traditional high-stake problems such as clinical trials, while experiments at online service companies typically have very different features and focuses. Motivated by the real needs, in this paper, we introduce a novel framework that we developed in Amazon to maximize customer experience and control opportunity cost. We formulate the problem as a Bayesian optimal sequential decision making problem that has a unified utility function. We discuss extensively practical design choices and considerations. We further introduce how to solve the optimal decision rule via Reinforcement Learning and scale the solution. We show the effectiveness of this novel approach compared with existing methods via a large-scale meta-analysis on experiments in Amazon.
胡竣峰
新加坡国立大学二年级博士,研究方向为时空数据挖掘,贝叶斯推断。研究成果发表在KDD, ECAI, TNNLS等会议和期刊上。
报告题目:Graph Neural Processes for Spatio-Temporal Extrapolation
张一帆
中国科学院自动化研究所直博二年级,研究兴趣:鲁棒/可靠的机器学习系统,研究成果已发表在ICLR,KDD,CVPR,ICML, TIP上,导师为谭铁牛院士。个人主页:https://yfzhang114.github.io/
报告题目:零成本测试时间自适应提升模型泛化性:域风险最小化
张凯科
中国科学院大学计算技术研究所硕博生二年级,研究方向为:可信图数据挖掘,鲁棒推荐系统。
报告题目:Disentangled Representation Learning for Discrete-time Dynamic Graph.
报告简介:Unsupervised representation learning for dynamic graphs has attracted a lot of research attention in recent years. Compared with static graph, the dynamic graph is a comprehensive embodiment of both the intrinsic stable characteristics of nodes and the time-related dynamic preference. However, existing methods generally mix these two types of information into a single representation space, which may lead to poor explanation, less robustness, and a limited ability when applied to different downstream tasks. To solve the above problems, in this paper, we propose a novel disentangled representation learning framework for discrete-time dynamic graphs. We specially design a temporal-clips contrastive learning task together with a structure contrastive learning to effectively identify the time-invariant and time-varying representations respectively. To further enhance the disentanglement of these two types of representation, we propose a disentanglement-aware discriminator under an adversarial learning framework from the perspective of information theory. Extensive experiments on Tencent and five commonly used public datasets demonstrate that DyTed, as a general framework that can be applied to existing methods, achieves state-of-the-art performance on various downstream tasks, as well as be more robust against noise.
往期精彩文章推荐
- 关注我们,记得星标 -
关于AI TIME
AI TIME源起于2019年,旨在发扬科学思辨精神,邀请各界人士对人工智能理论、算法和场景应用的本质问题进行探索,加强思想碰撞,链接全球AI学者、行业专家和爱好者,希望以辩论的形式,探讨人工智能和人类未来之间的矛盾,探索人工智能领域的未来。
迄今为止,AI TIME已经邀请了1100多位海内外讲者,举办了逾550场活动,超600万人次观看。
我知道你
在看
哦
~
点击 阅读原文 预约直播!