| Document Date: 2013-05-20 14:59:49
Open Document File Size: 174,55 KBShare Result on Facebook
City Cascais / Supercomputing / New York / Washington / DC / Asuncion / / Company LazyTables / Yahoo! LDA / HP / Microsoft / New york times / / Country Jordan / United States / Portugal / / Facility Eric Xing Carnegie Mellon University / / IndustryTerm update operator / convergent algorithms / distributed machine learning applications / staleness tolerant algorithms / iterative convergent algorithms / potential solution / Internet services / potential solutions / data mining / software components / parameter server / improved solution / staleness-tolerant algorithm / sparse regression algorithm / document topic modeling algorithm / problem solution / collaborative filtering algorithms / important applications / system solvers Web search/page rank Mapping / natural language processing / cloud computing / scientific computing / learning algorithms / / OperatingSystem VMs / / Organization USENIX Association / Carnegie Mellon University / IEEE Computer Society / / Person James Cipar / Kimberly Keeton / S. Pakin / Jin Kyu Kim / Gregory R. Ganger / D. J. Kerbyson / F. Petrini / Garth Gibson / / Position model for updating shared data / model for parallel computation / application programmer / / ProgrammingLanguage DC / C++ / / ProvinceOrState New York / California / / PublishedMedium Machine Learning / / Technology genomics / staleness-tolerant algorithms / RAM / bioinformatics / unix / API / collaborative filtering algorithms / LDA algorithm / Machine learning / staleness-tolerant algorithm / operating system / operating systems / iterative convergent algorithms / 8 / 192 processors / natural language processing / staleness tolerant algorithms / synchronous parallel These algorithms / example algorithms / data mining / sparse regression algorithm / document topic modeling algorithm / / URL http /
SocialTag |
Solving the straggler problem with bounded staleness James Cipar, Qirong Ho, Jin Kyu Kim, Seunghak Lee, Gregory R. Ganger, Garth Gibson, Kimberly Keeton*, Eric Xing Carnegie Mellon University, *HP Labs Algorithm