Publications

Journal articles / Preprints

  • Cartan moving frames and the data manifolds, to appear in Information Geometry, 2024, Co-author(s): Eliot Tron, Rita Fioresi, Stéphane Puechmorel.
  • Adversarial attacks on neural networks through canonical Riemannian foliations, to appear in Machine Learning, 2024, Co-author(s): Eliot Tron, Stéphane Puechmorel.
  • A Time-Dependent Subgraph-Capacity Model for Multiple Shortest Paths and Application to CO2/Contrail-Safe Aircraft Trajectories, Operations Research Forum, Volume 5 (71), 2024, Co-author(s): Céline Demouge, Marcel Mongeau, Daniel Delahaye
  • Climate-aware air traffic flow management optimization via column generation, Transportation Research Part C: Emerging Technologies, Volume 166, https://doi.org/10.1016/j.trc.2024.104792, 2024, Co-author(s): Céline Demouge, Marcel Mongeau, Daniel Delahaye.
  • Distributional loss for convolutional neural network regression and application to parameter estimation in satellite navigation signals, Expert Systems with Applications, in press https://doi.org/10.1016/j.eswa.2024.124682, 2024, Co-author(s): Thomas Gonzalez, Antoine Blais and Christian Ruiz.
  • MIQCQP reformulation of the ReLU neural networks Lipschitz constant estimation problem, eprint, arXiv:2402.01199, 2024, Co-author(s): Mohammed Sbihi and Sophie Jan.
  • Canonical foliations of neural networks: application to robustness, eprint, arXiv:2203.00922, 2022. Co-Author(s): Eliot Tron and Stéphane Puechmorel.
  • A Chernov bound for robust tolerance design and application, The International Journal of Advanced Manufacturing Technology , http://doi.org/10.1007/s00170-020-06231-8 , 2020. Co-author(s): Ambre Diet, Xavier Gendre and Julien Martin.
  • Convolutional neural network for multipath detection in GNSS receivers, eprint, arXiv:1911.02347, 2019. Co-author(s): Evegenii Munin and Antoine Blais.
  • The coupling effect of Lipschitz regularization in deep neural networks, to appear in SN Computer Science. 2021. eprint version: arXiv:1904.06253.
  • Using Wasserstein-2 regularization to ensure fair decisions with Neural-Network classifier, eprint, arXiv:1908.05783, 2019. Co-author(s):Laurent Risser, Quentin Vincenot, Jean-Michel Loubès.
  • On the Convergence of Stochastic Bi-level Gradient Methods, e-print Optimization Online. Co-author(s): W. Wang.
  • Self Adaptive Support Vector Machine: A Multi-Agent Optimization Perspective, Expert Systems With Applications, vol. 42, iss. 9, pp. 4284-4298, 2015. http://dx.doi.org/10.1016/j.eswa.2015.01.028. Co-author(s): S. Jan, T. Jorquera, J-P. Georgé.
  • Incremental Accelerated Gradient Methods for SVM Classification:Study of the Constrained A statistical approach for tolerancing from design stage to measurement analysisApproach, Computational Management Science, Special issue on learning and robustness, vol. 11, iss. 4, 2014. Co-author(s): S. Jan
  • An Adaptive Weighted Kernel Technique for Online Training with Imbalanced Data, submitted. Co-author(s): T.B. Trafalis
  • An Incremental Primal-Dual Method For Nonlinear Programming with Special Structure, Optimization Letters,  vol. 7, iss. 1, pp. 51-62, 2013. Co-author(s): T.B. Trafalis
  • Online SVM learning via an incremental primal-dual technique, Optimization Methods and Software, vol. 28, iss. 2, pp.256-275, 2013. Co-author(s): T.B. Trafalis
  • Neural network training via an affine scaling quadratic optimization algorithm,  Neural Networks, vol. 9, iss. 3, pp. 475-481, 1996. Co-author(s): T.B. Trafalis

Book chapters

  • Interior Point Methods for Supervised Training of Artificial Neural Networks with Bounded Weights , Pardalos, P. M., Hearn, D., and Hager, W., Eds.,New York, NY, USA: Springer, vol. 450, pp. 441-470, 1997, Co-author(s): T.B. Trafalis and T. A. Tutunji
  • Neural Network Training via Quadratic Programming , Barr, R. S., Helgason, R. V., and Kennington, J. L., Eds., Boston, MA, USA: Kluwer Academic Publishers, 1997, pp.123-139. Co-author(s): T.B. Trafalis

Published refereed conference proceedings

  • A statistical approach for tolerancing from design stage to measurement analysis, Proceedings of the 16th CIRP Conference on Computer Aided Tolerancing (e-Conference on June 15-17, 2020), Procedia CIRP. Co-author(s):  A. Diet, X. Gendre, J. Martin, J.-P. Navarro
  • Convolutional Neural Network for Multipath Detection in GNSS Receivers, Proceedings of the 2020 International Conference on Artificial Intelligence and Data Analytics for Air Transportation (AIDA-AT), IEEE, Singapore, Singapore, 2020, pp. 1-10, doi: 10.1109/AIDA-AT48540.2020.9049188. Co-author(s): E. Munin, A. Blais
  • On the use of generativze adversarial networks for aircraft trajectory generation and atypical approach detection, EIWAC 2019:, 6th ENRI International Workshop on ATM/CNS, Oct 2019, Tokyo, Japan. Co-author(s): G. Jarry, D. Delahaye
  • Feature uncertainty bounding schemes for large robust nonlinear SVM classifiers, Learning and Intelligent Optimization Conference, Proceedings of Learning and Intelligent Optimization Conference, LION’12, Kalamata, Greece, to appear in Lectures Notes in Computer Science, 2018. Co-author(s) : S. Jan.
  • An incremental nonlinear primal-dual algorithm and applications to artificial neural networks training, in Large scale systems: theory and applications 1998, Tarrytown, NY, USA, 1999, pp. 1003-1009. Co-author(s): T.B. Trafalis
  • An Affine Scaling Neural Network Training Algorithm for Prediction of Tornados, Proceedings of APMOD’98, Limassal, Cyprus, March 11-13, 1998. Co-author(s): T.B. Trafalis, P.I. Li, F. Stumpf and A. White
  • Training of supervised neural networks via a nonlinear primal-dual interior-point method, in Proceedings of the 1997 IEEE International Conference on Neural Networks, Piscataway, NJ, USA, 1997, pp. 2017-2021, 1997. Co-author(s): T. B. Trafalis and S. C. Bertrand
  • Affine scaling neural network training algorithm for prediction of tornados, in Intelligent Engineering Systems Through Artificial Neural Networks, New York, NY, USA, 1997, pp. 213-218, 1997. Co-author(s): T. B. Trafalis, P. -I. Li, G. Stumpf, and A. White
  • Training of Supervised Neural Networks via a Primal-Dual Interior Point Method for Nonlinear Programming, in Proceedings of the 10th Mid-America Symposium on Emerging Computer Technologies 1996, Norman, OK, USA, 1996, pp. 21-25. Co-author(s): T. B. Trafalis and S. C. Bertrand
  • Neural network training via a primal-dual interior point method for nonlinear programming, in World Congress on Neural Networks: International Neural Network Society 1996 Annual Meeting, Mahwah, NJ, USA, 1996, pp. 201-204. Co-author(s): T. B. Trafalis
  • Neural Network Training Via a Trust Region Algorithm for Quadratic Programming,  in Proceedings of the 9th Mid-America Symposium on Emerging Technologies 1995, Norman, OK, USA, 1995, pp. 23-28. Co-author(s): T. B. Trafalis
  • An incremental nonlinear primal-dual algorithm and applications to artificial neural networks training, in Large scale systems : theory and applications, 1995, Oxford, UK, 1995. Co-author(s): T. B. Trafalis
  • Neural network training via a primal-dual interior point method for linear programming, in World Congress on Neural Networks: 1994 International Neural Network Society Annual Meeting, Hillsdale, NJ, USA, 1994, pp. 798-803. Co-author(s): T. B. Trafalis

Presentation in conferences

  • Bi-level Stochastic Gradient for Large Scale Support Vector Machine, EUROPT 2014, 12th EUROPT Workshop in Continuous Optimization, Perpignan, 2014. Co-author: W. Wang
  • A Natural Formalism and a Multi-Agent Algorithm for Integrative Multidisciplinary Design Optimization, OPTMAS 2013. Co-author(s): T. Jorquera, J.P. Georgé, M.P. Gleizes, V. Noël and C. Régis
  • An Adaptive Weighted Kernel Technique for Online Training with Imbalanced Data, in EURO 2012, 25th European Conference on Operational Research, Vilnius, Lithuania, 2012. Co-author(s): T.B. Trafalis
  • An Incremental Interior Point Method for On-line SVM Learning, in AFG’11: 15th Austrian-French-German Conference on Optimization, Toulouse, France, 2011. Co-author(s): T.B. Trafalis
  • Online SVM learning via an incremental primal-dual technique, in WCGO 2011: The Second World Congress on Global Optimization, Chania, Greece, 2011. Co-author(s): T.B. Trafalis
  • A Stochastic Primal-Dual Technique for NonlinearProgramming, invited talk at EURO XVI, July 12-15, 1998. Co-author(s): T.B. Trafalis
  • An Incremental Primal-Dual Technique for Nonlinear Programming and Applications to Artificial Neural Network Training, International Symposium on Mathematical Programming, Lausanne, Switzerland, August 1997. Co-author(s): T.B. Trafalis
  • An Incremental Nonlinear Primal-Dual Algorithm for Neural Network Training and Applications to Financial Forecasting, invited talk at the Spring INFORMS/CORS 1998 meeting, Montreal, CANADA, April 1998. Co-author(s): T.B. Trafalis
  • Feedforward Supervised Neural Networks for Stock Price Prediction, invited talk at the Joint International Meeting EURO XV-INFORMS, Barcelona, Spain,
    July 1997. Co-author(s): T.B. Trafalis
  • Nonlinear Primal-Dual Methods for Neural Network Training and Applications, invited talk at INFORMS, Dallas, TX, Fall 1997. Co-author(s): T.B. Trafalis
  • Training of Supervised Neural Networks via a Primal-Dual Interior Point Method for Nonlinear Programming, MASECT ‘96. Co-author(s): T.B. Trafalis and S.C. Bertrand
  • Training of Artificial Neural Networks via Primal-Dual Interior Point Methods, invited talk at INFORMS meeting, Atlanta, GA, November 1996. Co-author(s): T.B. Trafalis
  • Interior Point Methods in Neural Network Training, INFORMS meeting, Washington, DC, May 1996. Co-author(s): T.B. Trafalis and T. Tutunji
  • Neural Network Training via Quadratic Programming, Fifth INFORMS Computer Science Technical Section Conference, Dallas, TX, January 8-10, 1996. Co-author(s): T.B. Trafalis
  • Application of Interior Point Methods to a Project Management Problem, ORSA/TIMS Spring 1994, Boston, April 1994. Co-author(s): T.B. Trafalis and S. Pulat