The conference theme, "Theory, Models, and Applications of Geometric Deep Learning," aligns with the current AI research focus on understanding AI algorithm capabilities and limitations. Deep learning methods, like convolutional neural networks and graph neural networks, have seen rich theoretical advancements. Geometric deep learning offers a unified framework, addressing properties such as permutation invariance and rotation invariance. Its applications span quantum computing, 3D perception, drug design, and more, exemplified by AlphaFold's protein structure predictions. However, these structure-aware networks often lack a solid mathematical foundation. This workshop unites mathematicians and computer scientists to establish the mathematical theory of geometric deep learning, fostering reliable topological structures and efficient computing units for deep neural networks while addressing current scientific challenges.
The Learning on Graphs conference (LoG) is an annual research conference that covers areas broadly related to machine learning on graphs and geometry, with a special focus on review quality. In its inaugural edition, LoG 2022 received 250+ paper submissions, 2,800+ total registrations, and distributed $30,000+ in reviewer awards.
LoG 2023 will be a 4-day virtual event via Zoom + GatherTown and completely free to attend (dates: 27th – 30th November 2023).
At the same time, LoG 2023 aims to host a ‘network’ of local mini-conferences around the world. This decision builds upon the success and positive reception of LoG 2022 local meetups in Cambridge, MIT, Stanford, Mila, and Würzburg (300+ attendees in total).
The objective is to connect participants belonging to the same geographic area, improving their social experience and fostering discussion and collaboration.
LoG 2023 Shanghai Meetup is held jointly by Shanghai Jiao Tong University and Peking University on 29 Nov - 1 Dec 2023.
TBA
The global conference is online with multiple local meetups around the world on 27 Nov – 1 Dec!. All the talks will be live-streamed on Zoom and YouTube with poster sessions and sponsor sessions on GatherTown: Schedule Zoom: https://logconference.org Youtube: https://youtube.com/@learningongraphs GatherTown: https://app.gather.town/app/etG6JBKcR2u5Qp4q/LOG%20Conference%202023
Our minisymposium is in hybrid format of in person talks and online talks. Zoom for online talks: For online participants, use the following link: Join from PC, Mac, Linux, iOS or Android: https://unsw.zoom.us/j/81596687439?pwd=enFLVjZXNkdVRFJmcG5keVJZejdLQT09 Meeting ID: 81596687439 Password: 613800 The conference will also be broadcast by WeChat Livestream. One can join by scanning the QR code below (also in the conference poster).
The first day (Nov 29) of our session is in Shanghai Minhang Platinum Hanjue Hotel.
The second and third days (Nov 30 - Dec 1) is in the Institute of Natural Sciences, Shanghai Jiao Tong University, which direction can refer to
https://ins.sjtu.edu.cn/articles/17
The speakers are invited, including keynotes, invited and lightning talks.
During the conference, there will be several poster presentations.
Registration deadline of the LoG 2023 Shanghai is Nov 29, 2023.
Please use the following eventbrite to register for the conference with your information (email, name, affiliation):
https://www.eventbrite.com/e/learning-on-graph-2023-tickets-729374477697?aff=oddtdtcreator
If you need an invitation letter for visa application, please do not hesitate to contact us.
All invited speakers will be provided accommodation in Shanghai Minhang Platinum Hanjue Hotel (上海闵行白金汉爵大酒店).
Other participants are recommended to book rooms in Platinum Hanjue Hotel early in advance.
For other hotels near the conference venue refer to ctrip.com, for example, Courtyard by Marriott Shanghai Minhang (紫竹万怡酒店).
The LoG 2023 Shanghai Meetup will be held on the Shanghai Nov 29 - Dec 1.
Keynote speakers are allotted 45 minutes and invited speakers 30 minutes including Q&A.
On Nov 30 - Dec 1, sessions will be held in Room 300 at Institute of Natural Science.
On Nov 29, sessions will be held in Platinum Hanjue Hotel.
This is PDF file to the conference program.
Zoom for online talks:
For online participants, use the following link:
Join from PC, Mac, Linux, iOS or Android:
https://unsw.zoom.us/j/81596687439?pwd=enFLVjZXNkdVRFJmcG5keVJZejdLQT09
Meeting ID: 81596687439
Password: 613800
Time Table |
||||
---|---|---|---|---|
Time | Session | Speaker | Talk Title | |
Day 1: 29 November, Wednesday | ||||
9:00-9:15 | Opening Remarks | |||
9:15-10:00 | Keynote talk | Zhangsheng Yu (SJTU) | Harnessing TME depicted by histological images to improve the cancer prognosis through a deep learning system | |
10:00-10:30 | Invited talk | Tailin Wu (Westlake) | GNN for scientific simulations: towards adaptive multi-resolution simulators and a foundation neural operator | |
10:30-11:00 | Morning tea | |||
11:30-12:00 | Invited talk | Yixin Liu (Monash) | Graph Neural Networks for Data with Less Information | |
11:30-12:00 | Invited talk | Yuanhong Jiang (SJTU) | Revisiting Scattering Transform with Message Passing Scheme | |
12:00-12:15 | Lightning talk | Jiale Yan (Tokyo Tech) | Multicoated and Folded Graph Neural Networks with Strong Lottery Tickets | |
12:15-14:00 | Lunch | |||
14:00-14:45 | Keynote talk | Junchi Yan (SJTU) | Machine Learning for Optimization Problems on Graphs | |
14:45-15:15 | Invited talk | Cheng Yang (BUPT) | Towards Graph Foundation Models | |
15:15-15:45 | Afternoon tea | |||
15:45-16:30 | Keynote talk | Stan Z. Li (Westlake) | AI for Life Sciences | |
16:30-17:30 | Panel discussion | |||
Day 2: 30 November, Thursday | ||||
9:00-9:10 | Shuttle to INS | |||
9:15-10:00 | Keynote talk | Xiaosheng Zhuang (CityU HK) | Framelets on Spheres and Graphs: Construction and Applications | |
10:00-10:30 | Invited talk | Eli Chien (Georgia Tech) | Graph, Differential Privacy, and Machine Unlearning | |
10:30-11:00 | Morning tea | |||
11:00-11:30 | Invited talk | Yiqing Shen (Johns Hopkins) | How GNNs Facilitate CNNs in Mining Geometric Information from Large-Scale Medical Images | |
11:30-12:00 | Invited talk | Ting Gao (HUST) | When stochastic dynamical systems meet the graph: long-term prediction and optimal control | |
12:00-14:00 | Lunch | |||
14:00-14:45 | Keynote talk | Xavier Bresson (NUS) | Graph Transformers and Developments | |
14:45-15:15 | Inivited talk | Minjie Wang (Amazon) | Graph Neural Network at Scale: A Tale of Productivity and Efficiency | |
15:15-15:45 | Afternoon tea | |||
16:00-16:30 | Invited talk | Dongmian Zou (Duke Kunshan) | All You Need is A Push: Generative Modeling via Gromov-Monge Embedding | |
16:30-17:00 | Invited talk | Cheng Cheng (SYSU) | Graph Filters and Distributed Algorithms on Spatially Distributed Networks | |
17:00-17:15 | Lightning talk | Chang Liu (Waterloo) | Kùzu: Graph Learning Applications Need a Modern Graph DBMS | |
18:30-21:00 | Dinner | |||
Day 3: 1 December, Friday | ||||
9:00-9:10 | Shuttle to INS | |||
9:15-10:00 | Keynote talk | Shirui Pan (Griffth) | Towards Unifying Large Language Models and Knowledge Graphs | |
10:00-10:30 | Invited talk | Sho Sonoda (RIKEN) | Deep Ridgelet Transform Induced from Group Invariant Functions | |
10:30-11:00 | Morning tea | |||
11:00-11:30 | Invited talk | Lu Bai (CUFE) | Transitive-Aligned Graph Neural Networks for Graph Classification | |
11:30-12:00 | Invited talk | Feng Xie (BTBU) | A "New" Separation Constraint and Its Application in Causal Discovery | |
12:00-12:15 | Lightning talk | Jing Gu (Duke Kunshan) | Three Revisits to Node-Level Graph Anomaly Detection: Outliers, Message Passing and Hyperbolic Neural Networks | |
12:15-12:30 | Lightning talk | Hao Chen (SJTU) | Lower and upper bounds for numbers of linear regions of graph convolutional networks | |
12:30-14:00 | Lunch | |||
14:00-14:45 | Keynote talk | Bohang Zhang (Peking) | Rethinking the Expressive Power of Graph Neural Networks: Insights from Distance, Biconnectivity, and Subgraph GNNs | |
14:45-15:15 | Inivited talk | Wei Duan (UTS) | Inferring Latent Temporal Sparse Coordination Graph for Multi-Agent Reinforcement Learning | |
15:15-15:45 | Afternoon tea | |||
15:45-16:15 | Invited talk | Xinliang Liu (KAUST) | Interpretable Neural Message Passing via Particle Systems and PDE Solvers | |
16:15-16:45 | Invited talk | Hoi To Wai (CUHK) | Learning Multiplex Graph with Inter-layer Coupling | |
16:45-17:15 | Invited talk | Kun Zhan (Lanzhou) | Graph Neural Estimators |