Parallel and Distributed Machine Learning

31 Views · 25 Downloads · ☆☆☆☆☆ 0.0

NOT PEER REVIEWED

Authors

Kashish Agarwal¹ , Aditya Bhat¹ , Atharva Deshpande¹ , Jayesh Bhave¹ , Fatima Inamdar¹

Poster summary: This poster outlines the difficulties, advantages, and distinctions between distributed and parallel machine learning (ML), navigating the complex field. It breaks down the difficulties in synchronisation, fault tolerance, and communication overhead that arise when parallelizing and spreading machine learning activities. On the other hand, it highlights the variety of advantages these methods provide, such as improved scalability, faster training rates, and the capacity to handle large datasets. Additionally, the poster clarifies the differences between distributed and parallel machine learning, explaining how distribution uses several machines in a network, whilst parallelism makes use of multiple resources on a single machine. The poster attempts to provide academics and practitioners with a comprehensive grasp of the intricacies surrounding parallel and distributed machine learning through lucid illustrations and succinct explanations, enabling well-informed decision-making.

Keywords: Parallel, Distributed, Data parallelism, Model paralellism, High Performance Computing

References

Parallel stochastic gradient descent for deep learning: A survey" by J. Dean et al. (2012)

Additional information

Data availability statement: Data sharing not applicable to this poster as no datasets were generated or analyzed during the current study.

Creative Commons license: Copyright © 2024 Agarwal et al. This is an open access work distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.