×

Hierarchical clustering with deep q-learning. (English) Zbl 1412.68182

Summary: Following up on our previous study on applying hierarchical clustering algorithms to high energy particle physics, this paper explores the possibilities to use deep learning to generate models capable of processing the clusterization themselves. The technique chosen for training is reinforcement learning, that allows the system to evolve based on interactions between the model and the underlying graph. The result is a model, that by learning on a modest dataset of 10, 000 nodes during 70 epochs can reach 83, 77% precision for hierarchical and 86, 33% for high energy jet physics datasets in predicting the appropriate clusters.

MSC:

68T05 Learning and adaptive systems in artificial intelligence
62H30 Classification and discrimination; cluster analysis (statistical aspects)
81-04 Software, source code, etc. for problems pertaining to quantum theory
81V05 Strong interaction, including quantum chromodynamics

Software:

SparseMatrix; Adam; Keras
PDFBibTeX XMLCite
Full Text: DOI arXiv

References:

[1] A. Ali, G. Kramer, Jets and QCD: A historical review of the discovery of the quark and gluon jets and its impact on QCD Eur. Phys. J. H. 36 (2011) 245-326. [arXiv:1012.2288 [hep-ph]]. ⇒ 89;
[2] D. Bader, J. McCloskey, Modularity and graph algorithms, SIAM AN10 Minisymposium on Analyzing Massive Real-World Graphs, 2009, pp. 12-16. ⇒ 92;
[3] J. W. Berry, B. Hendrickson, R. A. LaViolette, C. A. Phillips, Tolerating the community detection resolution limit with edge weighting, Phys. Rev. E83, 5 (2011) 056119. ⇒ 92;
[4] V. D. Blondel, J-L. Guillaume, R. Lambiotte, E. Lefebvre, Fast unfolding of communities in large networks, Journal of Statistical Mechanics: Theory and Experiment10 (2008) P10008 ⇒ 91; · Zbl 1459.91130
[5] S. Carani, Yu. L Dokshitzer, M. H. Seymour, B. R. Webher, Longitudinally-invariant \(k_⊥\)-clustering algorithms for hadron-hadron collisions, Nuclear Physics B406 (1993) 187-224. ⇒ 91;
[6] I. Csabai, F. Czakó, Z. Fodor, Quark- and gluon-jet separations using neural networks, Phys. Rev. D44 7 (1991) R1905-R1908. ⇒ 94;
[7] T. Davis, Y. Hu, The University of Florida Sparse Matrix Collection, Mathematical Software, Vol 38, Issue 1, 2011, pp 1:1-1:25. ⇒ 104; · Zbl 1365.65123
[8] B. Denby, Neural networks and cellular automata in experimental high energy physics, Computer Physics Communications49 (1988) 429-448. ⇒ 94;
[9] B. Denby, Neural networks in high energy physics: a ten year perspective, Computer Physics Communications119 (1999) 219. ⇒ 92;
[10] R. Forster, Louvain community detection with parallel heuristics on GPUs, 20th Jubilee IEEE International Conference on Intelligent Engineering Systems20 (2016) doi: 10.1109/INES.2016.7555126 ⇒ 91, 103;
[11] R. Forster, A. Fülöp, Jet browser model accelerated by GPUs, Acta Univ. Sapientiae Informatica8, 2 (2016) 171-185. ⇒ 91; · Zbl 1407.82006
[12] R. Forster, A. Fülöp, Parallel \(k_t\) jet clustering algorithm, Acta Univ. Sapientiae Informatica9, 1 (2017) 49-64. ⇒ 91; · Zbl 1371.65140
[13] R. Forster, A. Fülöp, Hierarchical \(k_t\) jet clustering for parallel achitectures, Acta Univ. Sapientiae Informatica9, 2 (2017) 195-213. ⇒ 87, 92; · Zbl 1383.81004
[14] X. Glorot, A. Bordes, Y-Bengio, Deep sparse rectifier neural networks, Proc 14th International Conference on Artificial Intelligence and Statistics (AISTATS) 2011, Fort Lauderdale, FL, USA. Volume 15 of JMLR:W&CP 15. ⇒ 101;
[15] J. Han. C. Moraga, The influence of the sigmoid function parameters on the speed of backpropagation learning, IWANN ’96 Proc. of the Int. Workshop on Artificial Neural Networks: From Natural to Artificial Neural Computation, 1995, pp. 195-201. ⇒ 102;
[16] R. Hahnloser, R. Sarpeshkar, M. A. Mahowald, R. J. Douglas, H. S. Seung, Digital selection and analogue amplification coexist in a cortex-inspired silicon circuit. Nature405 (2000) 947-951. ⇒ 101;
[17] D. P. Kingma, J. B. Adam, A method for stochastic optimization, 2014, arXiv:1412.6980 ⇒ 102;
[18] H. Kolanoski, Application of artifical neural networks in particle physics, Nuclear Instruments and Methods in Physics Research A367 (1995) 14-20. ⇒ 92;
[19] P. T. Komiske, E. M. Metodiev, M. D. Schwartz, Deep learning in color: towards automated quark/gluon jet discrimination, J. High Energy Physics (2017) 110. ⇒ 94; · Zbl 1373.81388
[20] K. J. C. Leney, A neural-network clusterisation algorithm for the ATLAS silicon pixel detector, J. of Physics: Conbnference Series523 (2014) 012023. ⇒ 94;
[21] L. Lönnblad, C. Peterson, T. Rögnvaldsson, Using neural networks to identify jets, Nuclear PhysicsB349 (1991) 675-702. ⇒ 94;
[22] H. Lu, Mahantesh Halappanavar, A. Kalyanaraman, Parallel heuristics for scalable community detection, Parallel Computing47 (2015) 1937. ⇒ 91;
[23] V. Mnih et al., Playing Atari with deep reinforcement learning, 2013, arXiv:1312.5602 ⇒ 99;
[24] V. Mnih et al., Human-level control through deep reinforcement learning, Nature, 2015, doi:10.1038/nature14236 ⇒ 99;
[25] T. Muta, Foundation of Quantum Chrodinamics, World Scientific Press, 1986. ⇒ 88;
[26] M. E. J. Newman, M. Girvan, Finding and evaluating community structure in networks, Phys. Rev. E69 2 (2004) 026113. ⇒ 92;
[27] M. E. Peskin, D. V. Schroeder, Quantum Field Theory, Westview Press, 1995. ⇒ 87;
[28] C. Peterson, Track finding with neural networks, Nuclear Instruments and MethodsA279 (1988) 537. ⇒ 94;
[29] D. Rohr, S. Gorbunov, A. Szostak, M. Kretz, T. Kollegger, T. Breitner, T. Alt, ALICE HLT TPC Tracking of Pb-Pb Events on GPUs, Journal of Physics: Conference Series396 (2012) doi:10.1088/1742-6596/396/1/012044 ⇒ 87, 104;
[30] G. P. Salam, Towards jetography, Eur. Phys. J. C67 (2010) 637-686. ⇒ 88;
[31] S. Salur, Full Jet reconstruction in heavy ion collisions, Nuclear Physics A830, 1-4 (2009) 139c-146c. ⇒ 89;
[32] K. E. Selbach, Neural network based cluster reconstruction in the ATLAS pixel detector, Nuclear Instruments and Methods in Physics Research A718 (2013) 363-365. ⇒ 94;
[33] N. Srivastava, G. Hinton, A. Krizhevsky, I. Sutskever, R. Salakhutdinov, Dropout: A simple way to prevent neural networks from overfitting, JMLR15 (2014) 1929-1958. ⇒ 102; · Zbl 1318.68153
[34] G. Sterman, S. Weinberg, Jets from quantum chromodynamics, Phys. Rev. Lett. 39 (1977) 1436. ⇒ 89, 90;
[35] R. S. Sutton, A. G. Barto, Reinforcement Learning: An Introduction, A Bradford Book, 1998, ISBN: 978-0262193986 ⇒ 94, 99;
[36] G. Swirszcz, W. M. Czarnecki, R. Pascanu, Local minima in training of neural networks, 2016, arXiv:1611.06310 ⇒ 100;
[37] V. A. Traag, P. Van Dooren, Y. Nesterov, Narrow scope for resolution-limit-free community detection, Phys. Rev. E84, 1 (2011) 016114. ⇒ 92;
[38] ∗ ∗ ∗ Keras: The Python Deep Learning library ⇒ 101;
[39] ∗ ∗ ∗ SuiteSparse Matrix Collection ⇒ 104;
[40] ∗ ∗ ∗ United States Census Bureau ⇒ 103;
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.