CNLS Postdoctoral Research Associate
Machine Learning in Geophysics
Office: TA-03, Bldg 1690, Room 107
Mail Stop: B216
Phone: (505) 000-0000
Fax: (505) 665-2659
- Convolutional auto-encoder and spectral analysis for predicting fault failure from seismic signals.
- Physics-informed convolutional auto-encoder for fluid flow in nano-cale pores of shale formation.
- Deep reinforcement learning for developing computational mechanics models.
- Multiscale LBM-DEM-FEM coupling for fractured porous media simulations.
- Ph.D. (2019) Computational Geomechanics, Columbia University, New York, USA
- M.S. (2015) Mechanical Engineering, Columbia University, New York, USA
- Dipl. Ing. (2013) Mechanical Engineering, University of Technology of Troyes, Troyes, France
- 2015-2019 Research Assistant, Department of Civil Engineering and Engineering Mechanics, Columbia University, New York, MY.
- Professional Training:
- 2019-present Postdoctoral Associate at Los Alamos National Laboratory, Los Alamos, NM.
- Multi-scale and multi-physics modeling of geological systems
- Physics-informed data-driven computational geomechanics.
- Reinforcement learning for material constitutive modeling.
Selected Recent Publications:
Google Scholar: Kun Wang
- K Wang, Y Chen, M Mehana, N Lubbers, KC Bennett, QJ Kang, HS Viswanathan, TC Germann, A physics-informed and hierarchically regularized data-driven model for predicting fluid flow through porous media, Journal of Computational Physics. under review, (2020).
- K Wang, KC Bennett, PA Johnson, Estimating fault slow slips from seismic signals via convolutional auto-encoder and spectral analysis, IEEE Transactions on Geoscience and Remote Sensing. under review, (2020).
- A Fuchs, Y Heider, K Wang, WC Sun, M Kaliske, DNN2: a hyperparameter reinforcement learning game for self-design neural network elasto-plastic constitutive laws, Computational Mechanics. under review, (2020).
- K Wang, WC Sun, Q Du, A non-cooperative meta-modeling game for automated third-party calibrating, validating and falsifying constitutive laws with parallelized adversarial attacks, Computer Methods in Applied Mechanics and Engineering. 373, (2020). Link
- Y Heider, K Wang, WC Sun, SO(3)-invariance of informed-graph-based deep neural network for anisotropic elastoplastic materials, Computer Methods in Applied Mechanics and Engineering. 363, (2020). Link