Estudo comparativo de plataformas de Deep Learning: Apache Singa, Graphlab e H2O

• Elias Augusto Fank UFFS
• Geomar A. Schreiner UFFS / UFSC
• Denio Duarte UFFS

DOI: https://doi.org/10.5753/erbd.2021.17234

Abstract

Deep learning techniques have been showing advances in various machine learning tasks. However, the implementation of these techniques is very complicated. Thus, to help with the implementation of Deep Learning projects, software platforms have been proposed. A considerable number of these platforms already exist, leading to difficulty choosing which platform is the better option to start a deep learning project. This work proposes a comparative study between some well-established platforms for Deep Learning projects: Apache Singa, Graphlab, and H20. This study aims to help users in choosing a platform. We conduct some experiments using two datasets: MNIST and KDD Cup 1999. Results show that the tested platforms are suitable for Deep Learning projects.

References

Bahrampour, S., Ramakrishnan, N., Schott, L., and Shah, M. (2015). Comparative study of deep learning software frameworks. In arXiv preprint arXiv:1511.06435.

Candel, A., Parmar, V., LeDell, E., and Arora, A. (2016). Deep learning with H2O. H2O.ai, Inc.

Deng, L. and Yu, D. (2014). Deep learning: methods and applications. Foundations and Trends in Signal Processing, 7(3–4):197–387.

Duarte, D. and Stahl, N. (2019). Machine learning: a concise overview. In Data Science in Practice, pages 27–58. Springer.

Guo, Y., Liu, Y., Oerlemans, A., Lao, S., Wu, S., and Lew, M. S. (2016). Deep learning for visual understanding: A review. Neurocomputing, 187:27–48.

Kovalev, V., Kalinovsky, A., and Kovalev, S. (2016). Deep learning with Theano, Torch, Caffe, TensorFlow, and Deeplearning4J: Which one is the best in speed and accuracy? In XIII International Conference on Pattern Recognition and Information Processing.

LeCun, Y., Bengio, Y., and Hinton, G. (2015). Deep learning. Nature, 521(7553):436–444.

Low, Y., Bickson, D., Gonzalez, J., Guestrin, C., Kyrola, A., and Hellerstein, J. M. (2012). Distributed GraphLab: a framework for machine learning and data mining in the cloud. Proceedings of the VLDB Endowment, 5(8):716–727.

Ng, S. S. Y., Zhu, W., Tang, W. W. S., Wan, L. C. H., and Wat, A. Y. W. (2016). An independent study of two deep learning platforms H2O and Singa. In 2016 IEEE International Conference on Industrial Engineering and Engineering Management (IEEM).

Ooi, B. C., Tan, K.-L., Wang, S., Wang, W., Cai, Q., Chen, G., Gao, J., Luo, Z., Tung, A. K. H., Wang, Y., Xie, Z., Zhang, M., and Zheng, K. (2015). SINGA: A distributed deep learning platform. In ACM Multimedia.

Shatnawi, A., Al-Bdour, G., Al-Qurran, R., and Al-Ayyoub, M. (2018). A comparative study of open source deep learning frameworks. In 9th ICICS, pages 72–77. IEEE.

Zimmerman, D. W. (1997). Teacher’s corner: A note on interpretation of the pairedsamples t test. Journal of Educational and Behavioral Statistics, 22(3):349–360.