Synthesizing of models for identification of teletraffic Markov chains by artificial neural networks and decision tree method

Balabanova, Ivelina Stefanova; Georgiev, Georgi Ivanov; Sadinov, Stanimir Michaylov; Kostadinova, Stela Savova

doi:10.2478/jee-2018-0055

Abstract

Imitation modelling processes of telegraphic systems on the Markov chains with unlimited and limited queues were made. For this purpose, the Java modeling tool simulation environment is used. With a fixed number of client stations and a number of system users, data are accumulated about the telegraphic system parameters as: customer ID, arrival time, server ID and exit system. Artificial neural networks (ANN) with backpropagation algorithm and decision tree (DT) method for identification of the studied Markov chains in MATLAB were applied. Training of the structural identification models to determine of the membership of the obtained parameters in telegraphic simulation to both unlimited and limited systems was carried out. The results of the training and synthesis of ANN and DT models are presented. Sufficient results have been obtained for telegraphic identification confirming the successful application of the proposed synthesized classification models, approximately 91% for DT and 99.2% for ANN.

References

[1] P. Sokólski and T. Rutkowski, “Hybrid of Neural Networks and Hidden Markov Models as a modern approach to speech recognition systems”, Pomiary Automatyka Robotyka, vol.17, no.2, pp.449-455, 2013.
Search in Google Scholar Back to article
[2] M. Mustafa, T. Allen and K. Appiah, “A comparative review of dynamic neural networks and hidden Markov model methods for mobile on-device speech recognition”, Springer, Neural Comput & Applic, pp.1-9, 2017.10.1007/s00521-017-3028-2
Search in Google Scholar Back to article
[3] M. Yang, W. Tu, W. Yin and Z. Lu, “Deep Markov Neural Network for Sequential Data Classification”, Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing, pp.32-37, 2015.10.3115/v1/P15-2006
Search in Google Scholar Back to article
[4] J. Bernstein, I. Dasgupta, D. Rolnick and H. Sompolinsky, “Markov Transitions between Attractor States in a Recurrent Neural Network”, AAAI Spring Symposium Series, Science of Intelligence: Computational Principles of Natural and Artificial Intelligence, pp.1-5, 2017.
Search in Google Scholar Back to article
[5] D. Lowd and J. Davis, “Learning Markov Network Structure with Decision Trees”, IEEE International Conference on Data Mining, Sydney, NSW, pp.334-343, 2010.10.1109/ICDM.2010.128
Search in Google Scholar Back to article
[6] D. Lowd and J. Davis, “Improving Markov Network Structure Learning Using Decision Trees”, Journal of Machine Learning Research (JMLR), no.15, pp.501-532, 2014.
Search in Google Scholar Back to article
[7] B. Lakshminarayanan, D. Roy and Y. The, “Top-Down Particle Filtering for Bayesian Decision Trees”, Proceedings of the 30th International Conference on Machine Learning, Atlanta, Georgia, USA, JMLR: W&CP, vol.28, pp.1-9, 2013.
Search in Google Scholar Back to article
[8] D. Bacciu, “Hidden Tree Markov Networks: Deep and Wide Learning for Structured Data”, Proceedings of the IEEE Symposium Series on Computational Intelligence (IEEE SSCI), Honolulu, HI, pp.1-8, 2017.10.1109/SSCI.2017.8285244
Search in Google Scholar Back to article
[9] E. Ivanova, T. Iliev, Gr. Mihaylov and R. Rashkov, “Significant Simulation Parameters for RESTART/LRE Method in Teletraffic Systems of Network of Queues”, IN: Automation Control Theory Perspectives in Intelligent Systems, Springer, pp.325-336, 2016.10.1007/978-3-319-33389-2_31
Search in Google Scholar Back to article
[10] E. Ivanova, T. Iliev, Gr. Mihaylov, I. Stoyanov, F. Tsvetanov, E. Otsetova and D. Radev, “Computer Modeling and Simulation of Teletraffic Models for 5G Networks”, 38th International Convention on Information and Telecommunication Technology, Electronics and Microelectronics-MIPRO, Croatia, pp.479-482, 2015.10.1109/MIPRO.2015.7160310
Search in Google Scholar Back to article

Synthesizing of models for identification of teletraffic Markov chains by artificial neural networks and decision tree method

Abstract

Paradigm

My account