{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,3,18]],"date-time":"2026-03-18T13:34:11Z","timestamp":1773840851770,"version":"3.50.1"},"reference-count":34,"publisher":"Association for Computing Machinery (ACM)","issue":"2","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Proc. VLDB Endow."],"published-print":{"date-parts":[[2018,10]]},"abstract":"<jats:p>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.<\/jats:p>","DOI":"10.14778\/3282495.3282499","type":"journal-article","created":{"date-parts":[[2019,1,4]],"date-time":"2019-01-04T13:35:28Z","timestamp":1546608928000},"page":"128-140","source":"Crossref","is-referenced-by-count":62,"title":["Rafiki"],"prefix":"10.14778","volume":"12","author":[{"given":"Wei","family":"Wang","sequence":"first","affiliation":[{"name":"National University of Singapore"}]},{"given":"Jinyang","family":"Gao","sequence":"additional","affiliation":[{"name":"National University of Singapore"}]},{"given":"Meihui","family":"Zhang","sequence":"additional","affiliation":[{"name":"University of Electronic Science and Technology of China"}]},{"given":"Sheng","family":"Wang","sequence":"additional","affiliation":[{"name":"National University of Singapore"}]},{"given":"Gang","family":"Chen","sequence":"additional","affiliation":[{"name":"Zhejiang University"}]},{"given":"Teck Khim","family":"Ng","sequence":"additional","affiliation":[{"name":"National University of Singapore"}]},{"given":"Beng Chin","family":"Ooi","sequence":"additional","affiliation":[{"name":"National University of Singapore"}]},{"given":"Jie","family":"Shao","sequence":"additional","affiliation":[{"name":"Beijing Institute of Technology"}]},{"given":"Moaz","family":"Reyad","sequence":"additional","affiliation":[{"name":"National University of Singapore"}]}],"member":"320","published-online":{"date-parts":[[2018,10]]},"reference":[{"key":"e_1_2_1_1_1","first-page":"265","volume-title":"OSDI 16","author":"Abadi M.","year":"2016","unstructured":"M. Abadi , P. Barham , J. Chen , Z. Chen , A. Davis , J. Dean , M. Devin , S. Ghemawat , G. Irving , M. Isard , M. Kudlur , J. Levenberg , R. Monga , S. Moore , D. G. Murray , B. Steiner , P. Tucker , V. Vasudevan , P. Warden , M. Wicke , Y. Yu , and X. Zheng . Tensorflow: A system for large-scale machine learning . In OSDI 16 , pages 265 -- 283 , GA, 2016 . USENIX Association. M. Abadi, P. Barham, J. Chen, Z. Chen, A. Davis, J. Dean, M. Devin, S. Ghemawat, G. Irving, M. Isard, M. Kudlur, J. Levenberg, R. Monga, S. Moore, D. G. Murray, B. Steiner, P. Tucker, V. Vasudevan, P. Warden, M. Wicke, Y. Yu, and X. Zheng. Tensorflow: A system for large-scale machine learning. In OSDI 16, pages 265--283, GA, 2016. USENIX Association."},{"key":"e_1_2_1_2_1","doi-asserted-by":"publisher","DOI":"10.5555\/2188385.2188395"},{"key":"e_1_2_1_3_1","volume-title":"Natural language processing (almost) from scratch. CoRR, abs\/1103.0398","author":"Collobert R.","year":"2011","unstructured":"R. Collobert , J. Weston , L. Bottou , M. Karlen , K. Kavukcuoglu , and P. P. Kuksa . Natural language processing (almost) from scratch. CoRR, abs\/1103.0398 , 2011 . R. Collobert, J. Weston, L. Bottou, M. Karlen, K. Kavukcuoglu, and P. P. Kuksa. Natural language processing (almost) from scratch. CoRR, abs\/1103.0398, 2011."},{"key":"e_1_2_1_4_1","first-page":"613","volume-title":"NSDI","author":"Crankshaw D.","year":"2017","unstructured":"D. Crankshaw , X. Wang , G. Zhou , M. J. Franklin , J. E. Gonzalez , and I. Stoica . Clipper: A low-latency online prediction serving system . In NSDI , pages 613 -- 627 , Boston, MA , 2017 . USENIX Association. D. Crankshaw, X. Wang, G. Zhou, M. J. Franklin, J. E. Gonzalez, and I. Stoica. Clipper: A low-latency online prediction serving system. In NSDI, pages 613--627, Boston, MA, 2017. USENIX Association."},{"key":"e_1_2_1_5_1","first-page":"235","volume-title":"CIDR","author":"Curino C.","year":"2011","unstructured":"C. Curino , E. Philip Charles Jones, R. Popa, N. Malviya, E. Wu, S. Madden, H. Balakrishnan, and N. Zeldovich. Relational cloud: A database-as-a-service for the cloud . CIDR , pages 235 -- 240 , April 2011 . C. Curino, E. Philip Charles Jones, R. Popa, N. Malviya, E. Wu, S. Madden, H. Balakrishnan, and N. Zeldovich. Relational cloud: A database-as-a-service for the cloud. CIDR, pages 235--240, April 2011."},{"key":"e_1_2_1_6_1","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2009.5206848"},{"key":"e_1_2_1_7_1","volume-title":"PANDA: facilitating usable AI development. CoRR, abs\/1804.09997","author":"Gao J.","year":"2018","unstructured":"J. Gao , W. Wang , M. Zhang , G. Chen , H. V. Jagadish , G. Li , T. K. Ng , B. C. Ooi , S. Wang , and J. Zhou . PANDA: facilitating usable AI development. CoRR, abs\/1804.09997 , 2018 . J. Gao, W. Wang, M. Zhang, G. Chen, H. V. Jagadish, G. Li, T. K. Ng, B. C. Ooi, S. Wang, and J. Zhou. PANDA: facilitating usable AI development. CoRR, abs\/1804.09997, 2018."},{"key":"e_1_2_1_8_1","volume-title":"Google Vizier: A Service for Black-Box Optimization","author":"Golovin D.","year":"2017","unstructured":"D. Golovin , B. Solnik , S. Moitra , G. Kochanski , J. E. Karro , and D. Sculley , editors . Google Vizier: A Service for Black-Box Optimization , 2017 . D. Golovin, B. Solnik, S. Moitra, G. Kochanski, J. E. Karro, and D. Sculley, editors. Google Vizier: A Service for Black-Box Optimization, 2017."},{"key":"e_1_2_1_9_1","doi-asserted-by":"publisher","DOI":"10.5555\/876875.879015"},{"key":"e_1_2_1_10_1","volume-title":"Deep residual learning for image recognition. CoRR, abs\/1512.03385","author":"He K.","year":"2015","unstructured":"K. He , X. Zhang , S. Ren , and J. Sun . Deep residual learning for image recognition. CoRR, abs\/1512.03385 , 2015 . K. He, X. Zhang, S. Ren, and J. Sun. Deep residual learning for image recognition. CoRR, abs\/1512.03385, 2015."},{"key":"e_1_2_1_11_1","volume-title":"Squeezenet: Alexnet-level accuracy with 50x fewer parameters and &lt;1 mb model size. CoRR, abs\/1602.07360","author":"Iandola F. N.","year":"2016","unstructured":"F. N. Iandola , M. W. Moskewicz , K. Ashraf , S. Han , W. J. Dally , and K. Keutzer . Squeezenet: Alexnet-level accuracy with 50x fewer parameters and &lt;1 mb model size. CoRR, abs\/1602.07360 , 2016 . F. N. Iandola, M. W. Moskewicz, K. Ashraf, S. Han, W. J. Dally, and K. Keutzer. Squeezenet: Alexnet-level accuracy with 50x fewer parameters and &lt;1 mb model size. CoRR, abs\/1602.07360, 2016."},{"key":"e_1_2_1_12_1","doi-asserted-by":"publisher","DOI":"10.14778\/3137628.3137664"},{"key":"e_1_2_1_13_1","first-page":"1106","volume-title":"NIPS","author":"Krizhevsky A.","year":"2012","unstructured":"A. Krizhevsky , I. Sutskever , and G. E. Hinton . Imagenet classification with deep convolutional neural networks . In NIPS , pages 1106 -- 1114 , 2012 . A. Krizhevsky, I. Sutskever, and G. E. Hinton. Imagenet classification with deep convolutional neural networks. In NIPS, pages 1106--1114, 2012."},{"key":"e_1_2_1_14_1","doi-asserted-by":"publisher","DOI":"10.1023\/A:1022859003006"},{"key":"e_1_2_1_15_1","doi-asserted-by":"publisher","DOI":"10.1038\/nature14539"},{"key":"e_1_2_1_16_1","doi-asserted-by":"publisher","DOI":"10.1145\/3187009.3177737"},{"key":"e_1_2_1_17_1","volume-title":"An overview. CoRR, abs\/1701.07274","author":"Li Y.","year":"2017","unstructured":"Y. Li . Deep reinforcement learning : An overview. CoRR, abs\/1701.07274 , 2017 . Y. Li. Deep reinforcement learning: An overview. CoRR, abs\/1701.07274, 2017."},{"key":"e_1_2_1_18_1","doi-asserted-by":"publisher","DOI":"10.14778\/2733004.2733081"},{"key":"e_1_2_1_19_1","volume-title":"Feb.","author":"Mnih V.","year":"2016","unstructured":"V. Mnih , A. Puigdom\u00e8nech Badia , M. Mirza , A. Graves , T. P. Lillicrap , T. Harley , D. Silver , and K. Kavukcuoglu . Asynchronous Methods for Deep Reinforcement Learning. ArXiv e-prints , Feb. 2016 . V. Mnih, A. Puigdom\u00e8nech Badia, M. Mirza, A. Graves, T. P. Lillicrap, T. Harley, D. Silver, and K. Kavukcuoglu. Asynchronous Methods for Deep Reinforcement Learning. ArXiv e-prints, Feb. 2016."},{"key":"e_1_2_1_20_1","first-page":"2017","article-title":"Tfx: A tensorflow-based production-scale machine learning platform","author":"Modi A. N.","year":"2017","unstructured":"A. N. Modi , C. Y. Koo , C. Y. Foo , C. Mewald , D. M. Baylor , E. Breck , H.-T. Cheng , J. Wilkiewicz , L. Koc , L. Lew , M. A. Zinkevich , M. Wicke , M. Ispir , N. Polyzotis , N. Fiedel , S. E. Haykal , S. Whang , S. Roy , S. Ramesh , V. Jain , X. Zhang , and Z. Haque . Tfx: A tensorflow-based production-scale machine learning platform . In KDD 2017 , 2017 . A. N. Modi, C. Y. Koo, C. Y. Foo, C. Mewald, D. M. Baylor, E. Breck, H.-T. Cheng, J. Wilkiewicz, L. Koc, L. Lew, M. A. Zinkevich, M. Wicke, M. Ispir, N. Polyzotis, N. Fiedel, S. E. Haykal, S. Whang, S. Roy, S. Ramesh, V. Jain, X. Zhang, and Z. Haque. Tfx: A tensorflow-based production-scale machine learning platform. In KDD 2017, 2017.","journal-title":"KDD"},{"key":"e_1_2_1_21_1","doi-asserted-by":"publisher","DOI":"10.1145\/2733373.2807410"},{"key":"e_1_2_1_22_1","volume-title":"Feb.","author":"Pham H.","year":"2018","unstructured":"H. Pham , M. Y. Guan , B. Zoph , Q. V. Le , and J. Dean . Efficient Neural Architecture Search via Parameter Sharing. ArXiv e-prints , Feb. 2018 . H. Pham, M. Y. Guan, B. Zoph, Q. V. Le, and J. Dean. Efficient Neural Architecture Search via Parameter Sharing. ArXiv e-prints, Feb. 2018."},{"key":"e_1_2_1_23_1","doi-asserted-by":"publisher","DOI":"10.1145\/2723372.2742911"},{"key":"e_1_2_1_24_1","volume-title":"Proximal policy optimization algorithms. CoRR, abs\/1707.06347","author":"Schulman J.","year":"2017","unstructured":"J. Schulman , F. Wolski , P. Dhariwal , A. Radford , and O. Klimov . Proximal policy optimization algorithms. CoRR, abs\/1707.06347 , 2017 . J. Schulman, F. Wolski, P. Dhariwal, A. Radford, and O. Klimov. Proximal policy optimization algorithms. CoRR, abs\/1707.06347, 2017."},{"key":"e_1_2_1_25_1","volume-title":"SE4ML: Software Engineering for Machine Learning (NIPS 2014 Workshop)","author":"Sculley D.","year":"2014","unstructured":"D. Sculley , G. Holt , D. Golovin , E. Davydov , T. Phillips , D. Ebner , V. Chaudhary , and M. Young . Machine learning: The high interest credit card of technical debt . In SE4ML: Software Engineering for Machine Learning (NIPS 2014 Workshop) , 2014 . D. Sculley, G. Holt, D. Golovin, E. Davydov, T. Phillips, D. Ebner, V. Chaudhary, and M. Young. Machine learning: The high interest credit card of technical debt. In SE4ML: Software Engineering for Machine Learning (NIPS 2014 Workshop), 2014."},{"key":"e_1_2_1_26_1","volume-title":"Very deep convolutional networks for large-scale image recognition. CoRR, abs\/1409.1556","author":"Simonyan K.","year":"2014","unstructured":"K. Simonyan and A. Zisserman . Very deep convolutional networks for large-scale image recognition. CoRR, abs\/1409.1556 , 2014 . K. Simonyan and A. Zisserman. Very deep convolutional networks for large-scale image recognition. CoRR, abs\/1409.1556, 2014."},{"key":"e_1_2_1_27_1","volume-title":"June","author":"Snoek J.","year":"2012","unstructured":"J. Snoek , H. Larochelle , and R. P. Adams . Practical Bayesian Optimization of Machine Learning Algorithms. ArXiv e-prints , June 2012 . J. Snoek, H. Larochelle, and R. P. Adams. Practical Bayesian Optimization of Machine Learning Algorithms. ArXiv e-prints, June 2012."},{"key":"e_1_2_1_28_1","volume-title":"Feb.","author":"Snoek J.","year":"2015","unstructured":"J. Snoek , O. Rippel , K. Swersky , R. Kiros , N. Satish , N. Sundaram , M. M. A. Patwary , Prabhat, and R. P. Adams . Scalable Bayesian Optimization Using Deep Neural Networks. ArXiv e-prints , Feb. 2015 . J. Snoek, O. Rippel, K. Swersky, R. Kiros, N. Satish, N. Sundaram, M. M. A. Patwary, Prabhat, and R. P. Adams. Scalable Bayesian Optimization Using Deep Neural Networks. ArXiv e-prints, Feb. 2015."},{"key":"e_1_2_1_29_1","first-page":"II-1139","volume-title":"ICML'13","author":"Sutskever I.","unstructured":"I. Sutskever , J. Martens , G. Dahl , and G. Hinton . On the importance of initialization and momentum in deep learning . ICML'13 , pages I II-1139 --III-1147. JMLR.org, 2013. I. Sutskever, J. Martens, G. Dahl, and G. Hinton. On the importance of initialization and momentum in deep learning. ICML'13, pages III-1139--III-1147. JMLR.org, 2013."},{"key":"e_1_2_1_30_1","doi-asserted-by":"publisher","DOI":"10.1109\/BigData.2017.8257923"},{"key":"e_1_2_1_31_1","doi-asserted-by":"publisher","DOI":"10.1007\/s00778-015-0391-4"},{"key":"e_1_2_1_32_1","doi-asserted-by":"publisher","DOI":"10.1145\/3003665.3003669"},{"key":"e_1_2_1_33_1","volume-title":"How good are machine learning clouds for binary classification with good features? CoRR, abs\/1707.09562","author":"Zhang H.","year":"2017","unstructured":"H. Zhang , L. Zeng , W. Wu , and C. Zhang . How good are machine learning clouds for binary classification with good features? CoRR, abs\/1707.09562 , 2017 . H. Zhang, L. Zeng, W. Wu, and C. Zhang. How good are machine learning clouds for binary classification with good features? CoRR, abs\/1707.09562, 2017."},{"key":"e_1_2_1_34_1","doi-asserted-by":"publisher","DOI":"10.1145\/3132847.3132944"}],"container-title":["Proceedings of the VLDB Endowment"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dl.acm.org\/doi\/pdf\/10.14778\/3282495.3282499","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2022,12,28]],"date-time":"2022-12-28T10:31:25Z","timestamp":1672223485000},"score":1,"resource":{"primary":{"URL":"https:\/\/dl.acm.org\/doi\/10.14778\/3282495.3282499"}},"subtitle":["machine learning as an analytics service system"],"short-title":[],"issued":{"date-parts":[[2018,10]]},"references-count":34,"journal-issue":{"issue":"2","published-print":{"date-parts":[[2018,10]]}},"alternative-id":["10.14778\/3282495.3282499"],"URL":"https:\/\/doi.org\/10.14778\/3282495.3282499","relation":{},"ISSN":["2150-8097"],"issn-type":[{"value":"2150-8097","type":"print"}],"subject":[],"published":{"date-parts":[[2018,10]]}}}