Neural Classification of HEP Experimental Data
High Energy Physics (HEP) experiments require discrimination of a few interesting events among a huge number of background events generated during an experiment. Hierarchical triggering hardware architectures are needed to perform this tasks in real-time. In this paper three neural network models are studied as possible candidate for such systems. A modified Multi-Layer Perceptron (MLP) architecture and a EαNet architecture are compared against a traditional MLP. Test error below 25% is archived by all architectures in two different simulation strategies. EαNet performance are 1 to 2%better on test error with respect to the other two architectures using the smaller network topology. The design of a digital implementation of the proposed neural network is also outlined.
KeywordsNeural Networks Intelligent Data Analysis Embedded Neural Networks
Unable to display preview. Download preview PDF.
- Beri, S.B., Bhat, P., Kaur, R., and Prosper, H. (2000). Top quark mass measurements using neural networks. Proc. of VII International Workshop on Advanced Computing and Analysis Techniques in Physics Research.Google Scholar
- Bhat, C. and Bhat, P. (2000). Using ensembles of neural networks in hep analysis. Proc. of VII International Workshop on Advanced Computing and Analysis Techniques in Physics Research.Google Scholar
- Ciobanu, C., Hughes, R., and Winer, B. (1999). Using neural networks to identify single top. CDF Note CDF/ANAL/TOP/GROUP/5370.Google Scholar
- Fent, J., Froechtenicht, W.,.Gaede, F, H. Getta, D. Goldner, and A. Gruber (1996). The realization of a second level neural network trigger for the h1 experiment at hera. Proc. of Fifth International Workshop On Software Engineering, Artificial Intelligence, Neural Nets, Genetic Algorithms, Symbolic Algebra, Automatic Calculation.Google Scholar
- Gaglio, S., Pilato, G., Sorbello, F., and Vassallo, G. (2000). Using the hermite regression formula to design a neural architecture with automatic learning of the hidden activation functions. Lecture Notes in Artificial Intelligence, Springer-Verlag, 1792:226–237.Google Scholar
- Hays, C. and Kotwal, A.V. (2002). Using a neural network for electron identification. CDF Note CDF/DOC/ELECTRON/CDFR/5810.Google Scholar
- Janauschek, L., Dichtl, J., Eberl, M., Enzenberger, M., and Fent, J. (1999). Artificial neural networks as a second level trigger at the h1 experiment at hera performance analysis and physics results. Proc. of Sixth International Workshop On Software Engineering, Artificial Intelligence, Neural Nets, Genetic Algorithms, Symbolic Algebra, Automatic Calculation.Google Scholar
- Pilato, G., Sorbello, F., and Vassallo, G. (2001). An innovative way to measure the quality of a neural network without the use of the test set. International Journal of Artificial Computational Intelligence, 5:31–36.Google Scholar
- Sorbello, F., Gioiello, G.A.M., and Vitabile, S. (1999). Handwritten character recognition using a mlp. L.C. Jain and B. Lazzerini Eds., Knowledge-Based Intelligent Techniques in Character Recognition, CRC Press, pages 91–119.Google Scholar
- Tuttle, J.P., Hays, C., and Kotwal, A.V. (2001). Neural networks for electron and photon identification. CDF Note CDF/DOC/ELECTRON/CDFR/5791.Google Scholar
- Vitabile, S., Gentile, A., and Sorbello, F. (2002). A neural network based automatic road signs recognizer. Proc. of IEEE World Congress on Computational Intelligence — International Joint Conference on Neural Networks, 3:2315–2320.Google Scholar
- Vitabile, S., Gentile, A., and Sorbello, F. (2004). Real-time road signs recognition on a simd architecture. WSEAS Transactions on Circuits and Systems, 3:664–669.Google Scholar