Gurpreet Singh

Contact adkbkk@r.postjobfree.com **** Greystone Drive,

Phone: +1-814-***-**** Austin, TX 78731.

Profile Google Scholar Researchgate Personal Website Linkedin Modeling

Skills

Machine Learning: SVM, DNN, CNN, AutoEncoders.

Algorithms: time-concurrency, parallel algorithms, distributed computing. Discretization schemes: nite di erence, nite volume, nite elements Subsurface modeling: multiphase ow, reactive transport, geomechanics. Oil & gas: water, gas, and chemical ooding, unconventional reservoirs Environmental: CO2 sequestration, groundwater remediation, contaminant migration. Software Skills Programming: Fortran, C/C++, Python, MPICH, OpenMP, OpenCL. Machine Learning: tensor ow, keras, scipy, scikit, pandas, pytorch, openCV Reservoir Simulator Development: IPARS, UTCOMP, PSim. Scienti c Toolsets: R, Octave, Matlab.

Education Ph.D., Petroleum Engineering, UT Austin. Fall 2014 Thesis: Coupled Flow and Geomechanics Modeling for Fractured Poroelastic Reservoirs Quantifying interaction between hydraulic and naturally occurring fractures is critical for evaluating production/injection and optimal well placement scenarios. I developed coupled fractured-reservoir ow and geomechanics model that provides better recovery estimates by capturing an accurate description of physics. M.S., Petroleum Engineering, Penn State. Fall 2008 Thesis: Finite Volume Analysis of Two Phase Flow Split at T-junctions Formation of gas condensate during natural gas transmission through pipeline networks drastically increases compression costs. I developed a two- uid nite volume model to predict condensed liquid pathways and suggest liquid removal scenarios to reduce pressure losses.

B. Tech., Chemical Engineering, GGS Indraprastha University Summer 2007 Delhi, India.

Professional

Experience

The University of Texas at Austin

Adjunct Faculty August 2019 to present

- Biomedical Engineering: Introduction to Numerical Methods

- Aerospace and Engineering Mechanics: Computational Engineering Computational Hydraulics Group, ICES

Research Scientist January 2019 to August 2020

- Developed a self correcting auto-encoder for HSI unmixing problem rely upon bi-orthogonal representations that outperforms SOTA results on benchmark problems.

- Developed a streaming singular value decomposition method that outperforms SOTA randomized SVD algorithms

- Developed a Hybrid Neural Pareto Filter (HNPF) approach for high-dimensional Pareto optimization problems to balance fairness and relevance in search results.

Center for Subsurface Modeling, ICES

Research Associate Mar 2017 to December 2018

1 of 2

- Developed a space-time domain decomposition approach with a time-concurrent, parallel solution algorithm for multiphase ow and reactive transport problems in subsurface porous media.

- Developed model order reduction techniques for practical problems of interest in subsurface porous media.

- Developed an approximate Jacobian non-linear solver that outperforms conventional preconditioners (CPR and two-stage).

- Project management and reporting for federal grants (DOE, NSF). Postdoctoral Research Fellow Dec 2014 to Mar 2017

- Developed adaptive homogenization approach for upscaling heterogeneous porous medium for computational e ciency and accuracy.

- Developed a computationally e cient and accurate fully implicit framework for subsurface reactive ow and transport processes.

- Extended the parallel framework in IPARS by modularizing MFMFE scheme to handle complex subsurface ow and transport processes.

- Assisted in planning and writing of federal grant proposals. Research Assistant May 2011 to Dec 2014

- Developed an EOS compositional ow model using advanced spatial discretization

(MFMFE).

- Developed and implemented coupled ow and geomechanics model for fractured poroelastic reservoirs.

Selected

Publications

1. Singh, G., Gupta, S., Lease, M. and Dawson C. \SCA-Net: A Self-correcting Two Layer Autoencoder for Hyperspectral Unmixing", arXiv:2102.05713, February 2021.

2. Singh, G., Gupta, S., Lease, M. and Dawson C. \A Hybrid 2-Stage Neural Optimization for Pareto Front Extraction", arXiv:2101.11684, February 2021. 3. Singh, G., Gupta, S., Lease, M. and Dawson C. \Streaming Singular Value Decomposition for Big Data Applications", arXiv:2010.14226, February 2021. 4. Singh, G., Gupta, S., Lease, M. and Dawson C. \TIME: A Transparent, Interpretable, Model-Adaptive and Explainable Neural Network for Dynamic Physical Processes", arXiv:2003.02426, March 2020. 5. Amanbek, Y., Singh, G. and Wheeler, M.F. \Adaptive Numerical Homogenization for Upscaling Single Phase Flow and Transport", Journal of Computational Physics, Vol. 387, pp. 117-133, June 2019.

6. Singh, G. and Wheeler, M.F. \A Space Time Domain Decomposition Approach using Enhanced VelocityMixed Finite Element Method", Journal of Computational Physics, Vol. 374, pp. 893-911, December 2018.

7. Singh, G., Pencheva, G., and Wheeler, M.F. \An Approximate Jacobian Nonlinear Solver for Multiphase Flow and Transport", Journal of Computational Physics, Vol. 375, pp. 337-351, December 2018.

8. Singh, G., Leung, W., and Wheeler, M.F. \Multiscale Methods for Model Order Reduction of Non Linear Multiphase Flow Problems", Computational Geosciences, November 2018.

9. Singh, G. and Wheeler, M.F. \Compositional Flow Modeling Using Multi- point Flux Mixed Finite Element Methods." Computational Geosciences, Vol. 20, no. 3, pp 421-435, June 2016.

2 of 2