Speedup Evaluation of the Training Algorithm for a Multilayer Perceptron based on Colored Petri Nets Simulations

• Matheus Dias Gama UFU
• Stéphane Julia UFU
• Rita Maria Silva Julia UFU

DOI: https://doi.org/10.5753/eniac.2021.18294

Resumo

Frequentemente, a complexidade da política de distribuição dos algoritmos distribuídos os tornam extremamente hostis a consagradas abordagens matemáticas de análise de desempenho, tais como análise assintótica, técnicas de recorrências e análise probabilística. Isso se deve ao fato de que tais métodos não provêm recursos adequados que lhes permitam avaliar o quão o gradual aumento do número de processadores impacta no tempo de execução do algoritmo. Diante disso, este artigo propõe uma abordagem visual e formal, baseada em simulações automáticas de modelos de Redes de Petri Coloridas Hierárquicas no ambiente gráfico Colored Petri Nets Tools (CPN Tools), para avaliar o speedup e o ponto de saturação de processadores de algoritmos distribuídos usados em Inteligência Artificial. Será usado como estudo de caso o algoritmo de treinamento dos Perceptrons de Múltiplas Camadas baseado na retro-propagação do erro.

Referências

Barbosa, V. C. (1996). An Introduction to Distributed Algorithms. The MIT Press, Cambridge, Massachusetts.

Borovska, P. and Lazarova, M. (2007). Efficiency of parallel minimax algorithm for game tree search. In Proceedings of the 2007 international conference on Computer systems and technologies, pages 1–6.

Brownlee, J. (2019). A gentle introduction to mini-batch gradient descent and how to configure batch size.

Cormen, T. H. (2009). Introduction to Algorithms. The MIT Press, Cambridge, Massachusetts.

Coulouris, G. F. (2012). Distributed systems: concepts and design/george coulouris...[et al.].

Goldsmith, S. F., Aiken, A. S., and Wilkerson, D. S. (2007). Measuring empirical comIn Proceedings of the the 6th joint meeting of the European putational complexity. software engineering conference and the ACM SIGSOFT symposium on The foundations of software engineering, pages 395–404.

Hanzálek, Z. (2003). Parallel Algorithms for Distributed Control-A Petri Net Based Approach. PhD thesis, Citeseer.

Hassoun, M. H. et al. (1995). Fundamentals of artificial neural networks. MIT press.

Hristev, R. (1998). The ann book.

Jensen, K., Christensen, S., Kristensen, L. M., and Westergaard, M. (2018). Cpn tools.

Jensen, K. and Kristensen, L. M. (2009). Coloured Petri nets: modelling and validation of concurrent systems. Springer Science & Business Media.

Júnior, C. M. M., Julia, R. M. S., Julia, S., and Silva, L. d. F. (2018). A new approach to evaluate the complexity function of algorithms based on simulations of hierarchical In Information Technology-New Generations, pages 555– colored petri net models. 564. Springer.

Kon, M. A. and Plaskota, L. (2000). Information complexity of neural networks. Neural Networks, 13(3):365–375.

Murata, T. (1989). Petri nets: Properties, analysis and applications. Proceedings of the IEEE, 77(4):541–580.

Sahu, M., Chakraborty, R., and Nayak, G. (2018). A task-level parallelism approach for process discovery. International Journal of Engineering & Technology, 7(4):2446–52.

Serpen, G. and Gao, Z. (2014). Complexity analysis of multilayer perceptron neural network embedded into a wireless sensor network. Procedia Computer Science, 36:192–197.