Statistical Learning Inference and Control (SLIC)

Quick overview

The Statistical Learning, Inference and Control (SLIC) project at HP Labs is focused on using statistical induction techniques and probabilistic models as a basis for automated decision making, management and control of complex IT systems.

One example of the research conducted by the SLIC project is the development of an automated performance diagnosis and forecasting engine for multi-tier systems. In this effort, probabilistic models are automatically generated, using pattern classification, that capture the relationships between relevant system utilization measurements and service level objective violations. These models are then use to locate probable causes of violations, identify normal regions of operation of the relevant measurements, and find early indicators of performance problems. Details can be found in our OSDI'04,  DSN'05, and SOSP'05 papers.

Team

• Ira Cohen
• Julie Symons
• Kivanc Ozonat
• Don Young (Project Manager)
• Raja Sambasivan (CMU, 2007 Intern)
• Diego Arias (UPRM, 2006-2007 Intern)
• Chengdu Huang (UIUC, 2006 Intern)
• Blaine Nelson (Berkeley, 2006 Intern)
• Nelson Lee (Stanford University, 2005 Intern)
• Rob Powers (Stanford University, 2004-2006 Intern)
• Steve Zhang (Stanford University, 2004-2006 Intern)
• Dawn Banard (Duke University, 2003 Summer Intern)

Collaborators

• Moises Goldszmidt, Microsoft Research
• Terence Kelly, HP Labs
• Kim Keeton, HP Labs
• Jeff Chase, Duke University
• George Forman, HP Labs
• Armando Fox, Stanford University
• Fabio Cozman, University of Sao Paulo

Related research

Learn more