Page 60 - Data Science Algorithms in a Week

44 Edwin Cortes, Luis Rabelo and Gene Lee

CONCLUSIONS

This research implemented a pattern recognition scheme to identify the best
optimistic time management and synchronization scheme to execute a particular Parallel
Discrete DES problem. This innovative pattern recognition approach utilizes Deep Belief
Neural Networks and measures of complexity to quantify and capture the structure of the
Parallel Discrete DES problem. This implementation of this approach was very
successful. That means that know we do not need to start doing by trial and error or
utilizing “inconsistent” and/or “fuzzy” rules in order to select the time management and
synchronization scheme. This method is direct (i.e., timeless execution) and selects
automatically the right scheme (i.e., TW, BTW, BTB).
A deep belief network model can be used as a detector of patterns not seeing during
training by inputting a mixture of diverse data from different problems in PDDES. In
reaction to the input, the ingested mixed data then triggers neuron activation probabilities
that propagate through the DBN layer-by-later until the DBN output is reached. The
output probability curve is then examined to select the best optimistic time management
and synchronization scheme (to be utilized).

REFERENCES

Bengio, Y. (2009). Learning deep architectures for AI. Foundations and trends in
Machine Learning, 2, 1-127.
Cho, K., Alexander, I. & Raiko, T. (2011). Improved learning of Gaussian-Bernoulli
restricted Boltzmann machines. In Artificial Neural Networks and Machine
Learning–ICANN 2011, 10-17.
Erhan, D., Yoshua, B., Courville, A., Manzagol, P., Pascal, V., & Bengio, S. (2010). Why
does unsupervised pre-training help deep learning? The Journal of Machine Learning
Research, 11, 625-660.
Fujimoto, R. (2000). Parallel and Distributed Simulation. New York: John Wiley &
Sons.
Hinton, G. (2007). Learning multiple layers of representation. Trends in cognitive
Sciences, 11(10), 428-434. doi:10.1016/j.tics.2007.09.004
Hinton, G. (2010). A practical guide to training restricted Boltzmann machines.
Momentum, 9(1), 926.
Hinton, G., Deng, L., Yu, D., Dahl, G., Mohamed, A., Jaitly, N., Senior, A., Vanhoucke,
V., Nguyen, P., Sainath, T., & Kingsbury, B. (2012). Deep neural networks for
acoustic modeling in speech recognition: The shared views of four research groups.
Signal Processing Magazine, IEEE, 29(6), 82-97. doi:10.1109/MSP.2012.2205597