Machine Learning and Systems

Expanding the applicability of AI/ML models under the constraints of real systems

About the lab

Machine Learning and Systems group at Nokia Bell Labs develops AI-based systems and fundamental algorithms that gap the bridge between the promise of limitless capabilities of AI and the constraints imposed by real systems. Our research addresses various aspects of such gaps along the chain of data-driven systems. On one hand, we address the limitations imposed by the data quality, heterogeneity, and lack of specifications, on the other hand, we develop ML-based models that leverage the human understanding and expertise of the engineered systems and comply with physical constraints of the execution environment such as network and computing capacity limitations. Finally, we aim at deriving AI-based end-to-end solutions with verifiable properties that ensure the trust towards the quality of the global system, provide explainability and causal interpretation of the outcomes and actions.