Combining Machine Learning and Control to Manage Computing System Complexity and Dynamics

Nikita Mishra; Connor Imes; John D. Lafferty; Henry Hoffmann. 16 January, 2019.
Communicated by Henry Hoffmann.


We present methods to simultaneously address the challenges of complexity and dynamics in computing systems. Our approach combines machine learning for modeling complex systems with control theory for adapting to dynamic events. The key is to develop a principled approach to the combination that keeps the strengths of both while covering the weaknesses.

This technical report is a summary of a longer research paper that appeared in ASPLOS 2018. This summary was submitted to IEEE Micro Top Picks, and received an honorable mention.

Original Document

The original document is available in PDF (uploaded 16 January, 2019 by Henry Hoffmann).