The wealth of digitized data forms the fundamental basis for the disruptive impact of Machine Learning. Yet a significant amount of data is scattered and locked in data silos, leaving its full potential untouched. Federated Machine Learning is a novel Machine Learning paradigm with the ability to overcome data silos by enabling the training of Machine Learning models on decentralized, potentially siloed data. Despite its advantages, most Federated Machine Learning projects fail in the project initiation phase due to their decentralized structure and incomprehensive interrelations. The current literature lacks a comprehensible overview of the complex project structure. Through a Design Science Research approach, we provide a process model of a Federated Machine Learning life cycle including required activities, roles, resources, artifacts, and interrelations. Thereby, we aim to aid practitioners in the project initiation phase by providing transparency and facilitating comprehensibility over the entire project life cycle.
Conference (0)
Name | Type | Size | Last Modification | Last Editor |
---|---|---|---|---|
230801 Mueller Process Model.pdf | 554 KB | 05.06.2023 |