Decentralized Machine Learning at the Edge

Many of today's parallel machine learning algorithms were developed for tightly coupled systems like computing clusters or clouds. However, the volumes of data generated from machine-to-machine interaction, by mobile phones or autonomous vehicles, surpass the amount of data that can be realistically centralized. Thus, traditional cloud computing approaches are rendered infeasible.

To scale parallel machine learning to such volumes of data, computation needs to be pushed towards the edge, that is, towards the data generating devices. By learning models directly on the data sources - which often have computational power of their own, for example, mobile phones, smart sensors, and tablets - network communication can be reduced by orders of magnitude. Moreover, it enables training a central model without centralizing privacy-sensitive data.

To foster discussion, discovery, and dissemination of novel ideas and approaches for decentralized machine learning, we organize a workshop at the European Conference on Machine Learning and Principles of Knowledge Discovery in Databases (ECMLPKDD) 2018 in Dublin. We welcome everyone to join the workshop and to submit their work. More information can be found on the website of DMLE'18.

Moreover, the workshop hosts two high-class invited talks. Dr. Michael May, head of the technology field Analytics & Monitoring at Siemens AG will speak about decentralized learning in the industrial context and Prof. Daniel Keren will give a talk on his latest research on decentralized and in-situ machine learning.