Abstract
In this paper we explore the scalability and distribution of a spatial data stream processing system by expanding a previously-proposed deep neural network-based classification system. Three distributions are studied: a centrally controlled cluster, a peer-to-peer system, and a hybrid of both. We also study the effect of expanding the number of classes being classified, including the gradual addition of classes to be classified to the neural network. We find that the deep neural network classification system does expand to more classes though it is not amenable to the gradual addition of classes during the networks training and that this is not improved by pre-training the network on all classes. We also find that the network performs asymmetrically depending on the number of classes being classified with optimal accuracy occurring with about half to a quarter as many classes as output nodes in the network. We finish this paper offering some recommendations to potentially use this system in a distributed manner to increase accuracy for the total set of classes.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
de Brébisson, A., Vincent, P.: An exploration of softmax alternatives belonging to the spherical loss family. In: Bengio, Y., LeCun, Y. (eds.) 4th International Conference on Learning Representations, ICLR 2016, San Juan, Puerto Rico, May 2-4, 2016, Conference Track Proceedings (2016)
Daghistani, A., Aref, W.G., Ghafoor, A.: TrioStat: online workload estimation in distributed spatial data streaming systems. In: Proceedings of the 28th International Conference on Advances in Geographic Information Systems, pp. 78–86 (2020)
Ding, Q., Ding, Q., Perrizo, W.: Decision tree classification of spatial data streams using peano count trees. In: Proceedings of the 2002 ACM Symposium on Applied Computing, pp. 413–417 (2002)
Gomes, H.M., Read, J., Bifet, A., Barddal, J.P., Gama, J.: Machine learning for streaming data: state of the art, challenges and opportunities. ACM SIGKDD Explor. Newlett. 21(2), 6–22 (2019)
Hershberger, J., Shrivastava, N., Suri, S.: Summarizing spatial data streams using clusterHulls. ACM J. Exp. Algorithms 13(2), 1–28 (2009)
King, L., Osborn, W.: An iterative strategy for deep learning classification on spatial data streams. In: Proceedings of the 19th International Conference on Advances in Mobile Computing and Multimedia, pp. 532–537 (2021)
Kingma, D.P., Ba, J.: Adam: a method for stochastic optimization. In: Bengio, Y., LeCun, Y. (eds.) Proceedings of the 3rd International Conference on Learning Representations (2015)
Masciari, E.: Finding homogeneous groups in trajectory streams. In: Proceedings of the 3rd ACM SIGSPATIAL International Workshop on GeoStreaming, pp. 11–18 (2012)
Mehta, P., Voisard, A., Mueller, S.: Clustering spatial data streams for targeted alerting in disaster response. In: Proceedings of the 4th ACM SIGSPATIAL International Workshop on GeoStreaming, pp. 66–75 (2013)
Nghiem, T.P., Waluyo, A.B., Taniar, D.: A pure peer-to-peer approach for \(k\)NN query processing in mobile ad hoc networks. Pers. Ubiquit. Comput. 17(5), 973–985 (2013)
Paszke, A., et al.: Pytorch: an imperative style, high-performance deep learning library. In: Advances in Neural Information Processing Systems (NeurIPS 2019), vol. 32, pp. 8024–8035 (2019)
Ryck, T.D., Lanthaler, S., Mishra, S.: On the approximation of functions by tanh neural networks. Neural Netw. 143, 732–750 (2021)
Sadashiv, N., Kumar, S.M.D.: Cluster, grid and cloud computing: a detailed comparison. In: Proceedings of the 6th International Conference on Computer Science and Education (ICCSE), pp. 477–482 (2011)
Sharifzadeh, M., Shahabi, C.: Approximate voronoi cell computation on spatial data streams. VLDB J. 18(1), 57–75 (2009)
Shekhar, S., Chawla, S.: Spatial Databases: A Tour. Prentice Hall, New Jersey (2003)
Yang, B., Garcia-Molina, H.: Comparing hybrid peer-to-peer systems. In: VLDB 2001, Proceedings of 27th International Conference on Very Large Data Bases, September 11-14, 2001, Roma, Italy, pp. 561–570 (2001)
Acknowledgements
We thank for reviewers of our paper for their constructive comments.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Clarke, K., Osborn, W. (2023). The Effects of Scale and Distribution on a Deep Neural Network Iterative Classification System of Spatial Data Streams. In: Barolli, L. (eds) Advances in Networked-based Information Systems. NBiS 2023. Lecture Notes on Data Engineering and Communications Technologies, vol 183. Springer, Cham. https://doi.org/10.1007/978-3-031-40978-3_3
Download citation
DOI: https://doi.org/10.1007/978-3-031-40978-3_3
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-40977-6
Online ISBN: 978-3-031-40978-3
eBook Packages: Intelligent Technologies and RoboticsIntelligent Technologies and Robotics (R0)