Bliss: auto-tuning complex applications using a pool of diverse lightweight learning models
Rohan Basu Roy, Tirthak Patel, Vijay Gadepally, Devesh Tiwari
Abstract
As parallel applications become more complex, auto-tuning becomes more desirable, challenging, and time-consuming. We propose, Bliss, a novel solution for auto-tuning parallel applications without requiring apriori information about applications, domain-specific knowledge, or instrumentation. Bliss demonstrates how to leverage a pool of Bayesian Optimization models to find the near-optimal parameter setting 1.64× faster than the state-of-the-art approaches.
Topics & Concepts
BLISSLeverage (statistics)Computer scienceA priori and a posterioriBayesian optimizationInstrumentation (computer programming)Domain (mathematical analysis)Artificial intelligenceMachine learningProgramming languageMathematical analysisEpistemologyMathematicsPhilosophyParallel Computing and Optimization TechniquesAdvanced Neural Network ApplicationsMachine Learning and Algorithms