MLbase: A Distributed Machine-learning System

By Tim Kraska, Ameet Talwalkar, John C. Duchi, Rean Griffith, M. Franklin, Michael I. Jordan. 2013

377 citations recorded by the source corpus.

How researchers see this paper

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

Citations in the corpus

Automated Machine Learning: State-of-The-Art and Open Challenges (2019, methodology): “VDS [107] has been recently introduced as an interactive automated machine learning tool, that followed the ideas of a previous work on the MLBase framework [67].”
Non-stochastic Best Arm Identification and Hyperparameter Optimization (2015, methodology): “We ignore this technicality in our experiments. such as LibSVM (Chang & Lin, 2011), scikit-learn (Pedregosa et al., 2011) and MLlib (Kraska et al., 2013).”
Towards an optimal stochastic alternating direction method of multipliers (2014, background): “also (Kraska et al., 2013)).”
SystemML: Declarative Machine Learning on Spark (2016, methodology): “Frameworks like MLBase [20, 32] or Columbus [42] further enable declarative ML tasks of model or feature selection, where they optimize both model accuracy and performance of ML algorithms.”
Understanding classifier errors by examining influential neighbors (2015, background): “This problem is growing in importance as systems allowing lay-users to train their own classifiers are emerging [27, 19, 20].”
Materialization optimizations for feature selection workloads (2014, background): “In a similar spirit, declarative machine learning systems, e.g., MLBase [26], provide the end users a high-level language to specify a machine learning task.”
The Missing Piece in Complex Analytics: Low Latency, Scalable Model Management and Serving with Velox (2014, methodology): “We could use a data-parallel modeling interface, such as MLbase [11] on Spark [21], to build a model of user preferences (say, in the form of a matrix representing predictions for user-item pairs) based on historical data—a batch-oriented task for which Spark is well suited and optimized.”
Democratizing Data Science through Interactive Curation of ML Pipelines (2019, background): “, classify X) has many commonalities with query optimization as already pointed out in the MLBase vision paper [19].”
Scaling Distributed Machine Learning with the Parameter Server (2014, unknown): “Our design de- Shared Data Consistency Fault Tolerance Graphlab [34] graph eventual checkpoint Petuum [12] hash table delay bound none REEF [10] array BSP checkpoint Naiad [37] (key,value) multiple checkpoint Mlbase [29] table BSP RDD Parameter (sparse) various continuous Server vector/matrix Table ”
TFX: A TensorFlow-Based Production-Scale Machine Learning Platform (2017, background): “Ideally, the platform automatically surveys diﬀerent machine learning techniques and suggests the best solution, allowing even non-experts access to machine learning [10].”
Smart Anomaly Detection in Sensor Systems (2020, methodology): “Distributed machine learning systems like MLbase [96] and Gaia [97] have been developed to run ML algorithms across different servers and over large-scale sensor systems.”
BayesDB: A probabilistic programming system for querying the probable implications of data (2015, methodology): “Other systems such as MLBase (Kraska et al., 2013) and GraphLab (Low et al., 2012) aim to simplify at-scale development and deployment of machine learning algorithms.”
PALM: Machine Learning Explanations For Iterative Debugging (2017, methodology): “While the data-base community has made substantial contributions in other areas of machine learning (e.g., training scalability and model speciﬁca-tion [1, 2, 5, 8, 10]), the opportunity is ripe for developing scalable tools to debug modern machine learning models.”
Proposal Scaling Distributed Machine Learning with System and Algorithm Co-design (2016, unknown): “This (key,value) abstraction is widely adopted by existing approaches [40, 75, 102].”
Towards Linear Algebra over Normalized Data (2016, unknown): “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 [1,9,10, 18,20,23,33].”