2017 · 44 citations · 35 references
ComputingComputer ScienceData ScienceBig Data AnalyticsDatabasesMachine LearningEnsemble LearningEnsemble MethodsMachine Learning For Big DataOptimization For Machine LearningScalable Machine LearningSentiment AnalysisDaily Intake PredictionDistributed Machine LearningInference EnginesKnowledge-based Expert SystemsMachine Learning AnalyticsMovement PredictionOnline Ensemble LearningOnline Machine Learning
Big data analytics is rapidly expanding, with machine learning models becoming essential for tasks such as sentiment analysis, food‑logging image classification, and stock prediction, yet extending traditional database systems to support these models is difficult due to implementation complexity and the need for expert optimization. This paper introduces Rafiki, a system that offers training and inference services for machine learning models. Rafiki delivers distributed hyper‑parameter tuning for training and online ensemble modeling for inference, balancing latency and accuracy. Experiments demonstrate Rafiki’s efficiency, effectiveness, scalability, and usability.
Big data analytics is gaining massive momentum in the last few years. Applying machine learning models to big data has become an implicit requirement or an expectation for most analysis tasks, especially on high-stakes applications. Typical applications include sentiment analysis against reviews for analyzing on-line products, image classification in food logging applications for monitoring user's daily intake, and stock movement prediction. Extending traditional database systems to support the above analysis is intriguing but challenging. First, it is almost impossible to implement all machine learning models in the database engines. Second, expert knowledge is required to optimize the training and inference procedures in terms of efficiency and effectiveness, which imposes heavy burden on the system users. In this paper, we develop and present a system, called Rafiki, to provide the training and inference service of machine learning models. Rafiki provides distributed hyper-parameter tuning for the training service, and online ensemble modeling for the inference service which trades off between latency and accuracy. Experimental results confirm the efficiency, effectiveness, scalability and usability of Rafiki.
35
Benchmarking cloud serving systems with YCSB
Brian F. Cooper, Adam Silberstein, Erwin Tam et al. · 2010 · 3.6K citations
Fay W. Chang, Sanjay Ghemawat, Wilson C. Hsieh et al. · ACM Transactions on Computer Systems · 2008 · 3.4K citations
Scalable SQL and NoSQL data stores
R. G. G. Cattell · ACM SIGMOD Record · 2011 · 1.3K citations