SystemML: Declarative machine learning on MapReduce

By Amol Ghoting, Rajasekar Krishnamurthy, Edwin P. D. Pednault, Berthold Reinwald, Vikas Sindhwani, Shirish Tatikonda, Yuanyuan Tian, Shivakumar Vaithyanathan. IEEE International Conference on Data Engineering, 2011

322 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition, Information Systems, Materials Chemistry

View Paper: https://www.semanticscholar.org/paper/d5b8b4bffc457150310dd3713241310a7f5c6165

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

In Information Systems

In Materials Chemistry

Citations in the corpus

KeystoneML: Optimizing Pipelines for Large-Scale Advanced Analytics (2016, methodology): “on plans and conventional optimizers do not consider such measures. Finally, the system should be aware of the computationvs-communication tradeoffs inherent in distributed processing of ML workloads [21, 34] and choose appropriate execution strategies in this regime. To address these challenges we ”
Towards Linear Algebra over Normalized Data (2016, methodology): “o 26x on real data. 1. INTRODUCTION The data management industry and academia are working intensively on tools to integrate machine learning (ML) algorithms and frameworks such as R with data systems [2,9,15, 16,18,20,31]. While almost all the ML tools require the input data to be presented as a sin”
Automating model search for large scale machine learning (2015, background): “These systems (all distributed with the exception of Weka), along with proprietary projects such as SystemML [21], all focus on training single models rather than model search.”
SystemML: Declarative Machine Learning on Spark (2016, methodology): “Example systems aiming for declarative ML algorithms are OptiML [34], SystemML [15], SimSQL [11], SciDB [33], Cumulon [18], DMac [39], and Mahout Samsara [24].”
Heterogeneity-aware Distributed Parameter Servers (2017, methodology): “The database community has developed a lot of ML systems, such as SystemML [20], SimSQL [10], Vertica-R [33], FactorizedLearning [26], and MADlib [22], in which ML analytics is expressed in a higher-level language and is compiled and executed in a MapReduce environment.”
MLbase: A Distributed Machine-learning System (2013, unknown): “Annals of Statistics, 28:2000, 1998.”
TuPAQ: An Efficient Planner for Large-scale Predictive Analytic Queries (2015, methodology): “These systems (all distributed with the exception of Weka), along with proprietary projects such as SystemML [26], all focus on training single models.”
Communication-Efficient Edge AI: Algorithms and Systems (2020, methodology): “Ghoting et al. [215] proposed SystemML based on the MapReduce framework to support distributed training for a broad class of supervised and unsupervised machine learning algorithms.”
Data Classification: Algorithms and Applications (2014, background): “It has been shown [44] that a declarative framework is particularly useful in many MapReduce applications, and that many existing classification algorithms can be generalized to the MapReduce framework.”
Probabilistic Demand Forecasting at Scale (2017, background): “SystemML [11] allows users to specify ML programs in a declarative, R-like language.”
Materialization optimizations for feature selection workloads (2014, methodology): “Our focus on performance optimizations across full programs was inspired by similar eﬀorts in RIOT-DB [32] and SystemML [17].”
Tupleware: Distributed Machine Learning on Small Clusters (2014, background): “Numerous extensions have been proposed to support iteration and shared state within MapReduce [10, 18, 7], and some projects (e.g., SystemML [19]) go a step further by providing a high-level language that is translated into MapReduce tasks.”