Videos & Recordings
All videos from the workshop have been uploaded here.
3rd Workshop on Distributed Machine Learning, co-located with CoNEXT 2022
Machine Learning and Deep Neural Networks are gaining more and more traction in a range of tasks such as image recognition, text mining as well as ASR. Commonly represented as execution graphs, they can be found deployed in a range of devices, from large servers in datacenters to small embedded devices, offering them intelligent capabilities. A common denominator between these diverse devices is that they are connected, thus enabling them to collaboratively work towards the learning task. Moreover, distributed ML can work as an enabler for various use-cases previously considered unattainable only using local resources.
Be it in a distributed environment, such as a datacenter, or a highly heterogeneous embedded deployment in the wild, distributed ML poses various challenges from a systems, interconnection and ML theoretical perspective. In this workshop, we want to focus on the network-side of such distributed setups, the unique challenges it presents and novel solutions for optimizing collaborative and distributed learning tasks. More specifically, we aim to welcome papers in the following areas:
- Distributed inference and offloading
- DNN computation sharing in local networks
- Distributed Learning
- Federated Learning
- Collaborative Learning
- Distributed and asynchronous training algorithms
- Channel optimisations for distributed ML
- DNN based compression schemes
- Fairness and biases in federated learning
- Security and privacy in distributed learning
- Interpretability in distributed/collaborative learning
- Novel ML applications in IoT, MEC, SDN or NFV scenarios
This year, we want to lay a special emphasis on two areas that we think are of specific interest in the area of DistributedML, i) that of robustness in learning under the presence of adversaries, as well as the ii) sustainability aspect of training in a distributed manner. We believe that these are key aspects becoming increasingly important to solve for successful and trustworthy distributed deployments in the wild.
We hope that DistributedML will serve as a forum for researchers across different disciplines to bring forward and discuss challenging topics, share new ideas and exchange experience in the deployment of such systems, both from a theoretical and experimental perspective.
The workshop is co-located with CoNEXT will be held in the beautiful city of Rome, in a hybrid manner (physically and virtually). All accepted paper will also be included in the conference proceedings and be made available in the ACM Digital Library.
The workshop will take place on the 9th of December 2022.
Deadline for paper submissions: 16th September 2022 23rd September 2022
