Matches in SemOpenAlex for { <https://semopenalex.org/work/W3029379969> ?p ?o ?g. }
- W3029379969 abstract "Distributed machine learning (ML) has triggered tremendous research interest in recent years. Stochastic gradient descent (SGD) is one of the most popular algorithms for training ML models, and has been implemented in almost all distributed ML systems, such as Spark MLlib, Petuum, MXNet, and TensorFlow. However, current implementations often incur huge communication and memory overheads when it comes to large models. One important reason for this inefficiency is the row-oriented scheme (RowSGD) that existing systems use to partition the training data, which forces them to adopt a centralized model management strategy that leads to vast amount of data exchange over the network. We propose a novel, column-oriented scheme (ColumnSGD) that partitions training data by columns rather than by rows. As a result, ML model can be partitioned by columns as well, leading to a distributed configuration where individual data and model partitions can be collocated on the same machine. Following this locality property, we develop a simple yet powerful computation framework that significantly reduces communication overheads and memory footprints compared to RowSGD, for large-scale ML models such as generalized linear models (GLMs) and factorization machines (FMs). We implement ColumnSGD on top of Apache Spark, and study its performance both analytically and experimentally. Experimental results on both public and real-world datasets show that ColumnSGD is up to 930× faster than MLlib, 63× faster than Petuum, and 14× faster than MXNet." @default.
- W3029379969 created "2020-06-05" @default.
- W3029379969 creator A5033860263 @default.
- W3029379969 creator A5040157606 @default.
- W3029379969 creator A5041232755 @default.
- W3029379969 creator A5057233052 @default.
- W3029379969 creator A5062357883 @default.
- W3029379969 creator A5081651393 @default.
- W3029379969 date "2020-04-01" @default.
- W3029379969 modified "2023-09-23" @default.
- W3029379969 title "C olumnSGD: A Column-oriented Framework for Distributed Stochastic Gradient Descent" @default.
- W3029379969 cites W1551558821 @default.
- W3029379969 cites W1845160311 @default.
- W3029379969 cites W1969048569 @default.
- W3029379969 cites W1980147176 @default.
- W3029379969 cites W1982767656 @default.
- W3029379969 cites W2005112390 @default.
- W3029379969 cites W2041517243 @default.
- W3029379969 cites W2110104287 @default.
- W3029379969 cites W2119738171 @default.
- W3029379969 cites W2146502635 @default.
- W3029379969 cites W2186615578 @default.
- W3029379969 cites W2247380138 @default.
- W3029379969 cites W2295739661 @default.
- W3029379969 cites W2460815216 @default.
- W3029379969 cites W2523246573 @default.
- W3029379969 cites W2579247884 @default.
- W3029379969 cites W2589642470 @default.
- W3029379969 cites W2741736887 @default.
- W3029379969 cites W2744829609 @default.
- W3029379969 cites W2751219672 @default.
- W3029379969 cites W2755021080 @default.
- W3029379969 cites W2945003059 @default.
- W3029379969 cites W2953812218 @default.
- W3029379969 cites W2954548380 @default.
- W3029379969 cites W2961280047 @default.
- W3029379969 cites W2963288913 @default.
- W3029379969 cites W2963390885 @default.
- W3029379969 cites W2964121744 @default.
- W3029379969 cites W2980342181 @default.
- W3029379969 doi "https://doi.org/10.1109/icde48307.2020.00134" @default.
- W3029379969 hasPublicationYear "2020" @default.
- W3029379969 type Work @default.
- W3029379969 sameAs 3029379969 @default.
- W3029379969 citedByCount "3" @default.
- W3029379969 countsByYear W30293799692021 @default.
- W3029379969 crossrefType "proceedings-article" @default.
- W3029379969 hasAuthorship W3029379969A5033860263 @default.
- W3029379969 hasAuthorship W3029379969A5040157606 @default.
- W3029379969 hasAuthorship W3029379969A5041232755 @default.
- W3029379969 hasAuthorship W3029379969A5057233052 @default.
- W3029379969 hasAuthorship W3029379969A5062357883 @default.
- W3029379969 hasAuthorship W3029379969A5081651393 @default.
- W3029379969 hasConcept C114614502 @default.
- W3029379969 hasConcept C120314980 @default.
- W3029379969 hasConcept C124101348 @default.
- W3029379969 hasConcept C126042441 @default.
- W3029379969 hasConcept C133875982 @default.
- W3029379969 hasConcept C135598885 @default.
- W3029379969 hasConcept C138885662 @default.
- W3029379969 hasConcept C154945302 @default.
- W3029379969 hasConcept C173608175 @default.
- W3029379969 hasConcept C199360897 @default.
- W3029379969 hasConcept C206688291 @default.
- W3029379969 hasConcept C26713055 @default.
- W3029379969 hasConcept C2779808786 @default.
- W3029379969 hasConcept C2780551164 @default.
- W3029379969 hasConcept C2781215313 @default.
- W3029379969 hasConcept C33923547 @default.
- W3029379969 hasConcept C41008148 @default.
- W3029379969 hasConcept C41895202 @default.
- W3029379969 hasConcept C42812 @default.
- W3029379969 hasConcept C50644808 @default.
- W3029379969 hasConcept C75684735 @default.
- W3029379969 hasConcept C76155785 @default.
- W3029379969 hasConcept C77088390 @default.
- W3029379969 hasConcept C91481028 @default.
- W3029379969 hasConceptScore W3029379969C114614502 @default.
- W3029379969 hasConceptScore W3029379969C120314980 @default.
- W3029379969 hasConceptScore W3029379969C124101348 @default.
- W3029379969 hasConceptScore W3029379969C126042441 @default.
- W3029379969 hasConceptScore W3029379969C133875982 @default.
- W3029379969 hasConceptScore W3029379969C135598885 @default.
- W3029379969 hasConceptScore W3029379969C138885662 @default.
- W3029379969 hasConceptScore W3029379969C154945302 @default.
- W3029379969 hasConceptScore W3029379969C173608175 @default.
- W3029379969 hasConceptScore W3029379969C199360897 @default.
- W3029379969 hasConceptScore W3029379969C206688291 @default.
- W3029379969 hasConceptScore W3029379969C26713055 @default.
- W3029379969 hasConceptScore W3029379969C2779808786 @default.
- W3029379969 hasConceptScore W3029379969C2780551164 @default.
- W3029379969 hasConceptScore W3029379969C2781215313 @default.
- W3029379969 hasConceptScore W3029379969C33923547 @default.
- W3029379969 hasConceptScore W3029379969C41008148 @default.
- W3029379969 hasConceptScore W3029379969C41895202 @default.
- W3029379969 hasConceptScore W3029379969C42812 @default.
- W3029379969 hasConceptScore W3029379969C50644808 @default.
- W3029379969 hasConceptScore W3029379969C75684735 @default.
- W3029379969 hasConceptScore W3029379969C76155785 @default.
- W3029379969 hasConceptScore W3029379969C77088390 @default.