Program
UTC 9:00-11:45 (EST 5:00-7:45, CEST 11:00-13:45, JST 18:00-20:45)
9:00-9:30 Schema Mapping between Logical and Internal Layers for NoSQL Applications [paper][recording]
Teruyoshi Zenmyo (CyberAgent, Inc.), Noriyuki Watanabe (CyberAgent, Inc.), Yusuke Wakuta (Osaka University), and Makoto Onizuka (Osaka University)
9:30-10:00 Yosegi: Columnar format for efficient nested data processing by schema conversion [paper][recording]
Yasunori Oto (Yahoo! JAPAN Corporation), Kouji Ijima (Yahoo! JAPAN Corporation), and Kotaro Terada (Yahoo Japan Corporation), Makoto Onizuka (Osaka University)
10:00-10:30 Exploiting SIMD Instructions in Partial-Evaluation-Based Query Compiler [paper][recording]
Jun Nemoto (Keio University), Hideyuki Kawashima (Keio University), and Motomichi Toyama (Keio University)
10:30-10:45 Break
10:45-11:45 Keynote speech: Distributed Deep Learning: Progress and Challenges [recording]
Dr. Takuya Akiba (Preferred Networks, Inc.)
UTC 15:00-17:45 (EST 11:00-13:45, CEST 17:00-19:45, JST 24:00-26:45)
15:00-16:00 Keynote speech: Shared Clusters for Machine Learning: Through the looking glass [recording part 1 | recording part 2] [slide]
Prof. Shivaram Venkataraman (University of Wisconsin, Madison)
16:00-16:15 Break
16:15-16:45 Masha: Sampling-Based Performance Prediction of Big Data Applications in Resource-Constrained Clusters [paper][recording]
Hani Al-Sayeh (TU-Ilmenau), Bunjamin Memishi (German Aerospace Center), Marcus Paradies (German Aerospace Center), and Kai-Uwe Sattler (TU Ilmenau)
16:45-17:15 Towards Low-Latency Automated Machine Learning [paper][recording]
Thomas Parnell (IBM Research), Dimitrios Sarigiannis (IBM Research), Andreea Anghel (IBM Research), and Haralampos Pozidis (IBM Research Zurich)
17:15-17:45 Analysis and Exploitation of Dynamic Pricing in the Public Cloud for ML Training [paper][recording]
Deepak Narayanan (Stanford), Keshav Santhanam (Stanford), Fiodar Kazhamiaka (Stanford), Amar Phanishayee (Microsoft Research), and Matei Zaharia (Stanford and Databricks)