Ivan Papusha

• Email: ivan.papusha@gmail.com
• Contact: Resume, LinkedIn, Github, Google Scholar
• Teaching: CDS 270-2

Bio

Ivan Papusha is a computational scientist and engineer. He is currently a senior analyst and subject matter expert in Washington, D.C., with over a decade of research and development experience in autonomy, air and space systems, and artificial intelligence. His (continuing) mission: to seek computational craftsmanship—the virtuous combination of mathematical insight, computing, and experimentation that expands to elegant solutions of engineering problems, and delivers rich new ways to think about the world.

He has a PhD from Caltech, and BS and MS degrees from Stanford University. He is a recipient of the NDSEG Fellowship, a member of IEEE, and the founder of Subgradient.

Publications

Autonomy and Machine Learning

• Y. Kouskoulas, T. J. Machado, D. Genin, A. Schmidt, I. Papusha, and J. Brulé. “Envelopes and Waves: Safe Multivehicle Collision Avoidance for Horizontal Non-deterministic Turns,” Journal on Software Tools for Technology Transfer, 2022. [arXiv] [doi] [Coq proofs]
• D. Genin, I. Papusha, J. Brulé, T. Young, G. Mullins, Y. Kouskoulas, R. Wu, and A. Schmidt. “Formal Verification of Neural Network Controllers for Collision-Free Flight,” Workshop on Numerical Software Verification, part of International Conference on Computer-Aided Verification (CAV), Los Angeles, CA, July 18–19, 2021. [doi] [NSV21]
• Y. Kouskoulas, D. Genin, A. Schmidt, I. Papusha, R. Wu, G. Mullins, T. Young, and J. Brulé. “Verification of Safety in Artificial Intelligence and Reinforcement Learning Systems,” Johns Hopkins APL Technical Digest, Volume 35, Number 4, 2021. [pdf]
• I. Papusha, R. Wu, J. Brulé, Y. Kouskoulas, D. Genin, and A. Schmidt. “Incorrect by Construction: Fine Tuning Neural Networks for Guaranteed Performance on Finite Sets of Examples,” Workshop on Formal Methods for ML-Enabled Autonomous Systems (FoMLAS), part of International Conference on Computer-Aided Verification (CAV), Los Angeles, CA, July 19, 2020. [arXiv] [lantern-smt code]

Air and Space Systems

• M. A. Kelly, J. L. Carr, D. L. Wu, A. C. Goldberg, I. Papusha, and R. T. Meinhold. “Compact Midwave Imaging System: Results from an Airborne Demonstration,” Remote Sensing, 14(4):834, 2022. [doi] [preprint]
• M. A. Kelly, D. Wu, A. Goldberg, I. Papusha, J. Wilson, J. Carr, J. Boldt, J. Greenberg, F. Morgan, S. Yee, A. Heidinger, and L. Mehr. “Compact Midwave Imaging System (CMIS) for Retrieval of Cloud Motion Vectors (CMVs) and Cloud Geometric Heights (CGHs),” Proceedings of the SPIE, Volume 10776, Remote Sensing of the Atmosphere, Clouds, and Precipitation VII, Honolulu, HI, September 24–26, 2018. [doi] [AGU19] [ESTO19] [slides]
• M. Cubuktepe, N. Jansen, S. Junges, J.-P. Katoen, I. Papusha, H. A. Poonawala, and U. Topcu. “Sequential Convex Programming for the Efficient Verification of Parametric MDPs,” International Conference on Tools and Algorithms for the Construction and Analysis of Systems (TACAS), pp. 133–150, Uppsala, Sweden, April 22–29, 2017. [arXiv] [doi]
• S. S. Farahani, I. Papusha, C. McGhan, and R. M. Murray. “Constrained Autonomous Satellite Docking via Differential Flatness and Model Predictive Control,” IEEE Conference on Decision and Control (CDC), pp. 3306–3311, Las Vegas, NV, December 12–14, 2016. [pdf] [doi]

Control Theory and Convex Optimization

• I. Papusha, U. Topcu, S. Carr, and N. Lauffer. “Affine Multiplexing Networks: System Analysis, Learning, and Computation,” arXiv:1805.00164 [math.OC], 2018. [arXiv] [slides] [amnet code]
• I. Papusha, M. Wen, and U. Topcu. “Inverse Optimal Control with Regular Language Specifications,” American Control Conference (ACC), pp. 770–777, Milwaukee, WI, June 27–29, 2018. [pdf] [doi] [slides]
• M. Wen, I. Papusha, and U. Topcu. “Learning from Demonstrations with High-Level Side Information,” International Joint Conference on Artificial Intelligence (IJCAI), pp. 3055–3061, Melbourne, Australia, August 19–25, 2017. [pdf] [doi]
• J. Fu, I. Papusha, and U. Topcu. “Sampling-based Approximate Optimal Control Under Temporal Logic Constraints,” ACM International Conference on Hybrid Systems: Computation and Control (HSCC), pp. 227–235, Pittsburgh, PA, April 18–20, 2017. [pdf] [doi] [slides]
• I. Papusha, J. Fu, U. Topcu, and R. M. Murray. “Automata Theory Meets Approximate Dynamic Programming: Optimal Control with Temporal Logic Constraints,” IEEE Conference on Decision and Control (CDC), pp. 434–440, Las Vegas, NV, December 12–14, 2016. [pdf] [doi] [slides] [sydar code]
• I. Papusha. Robustness, Adaptation, and Learning in Optimal Control. Ph.D. thesis, California Institute of Technology, 2016. [pdf] [CaltechTHESIS] [doi]
• I. Papusha and R. M. Murray. “Analysis of Control Systems on Symmetric Cones,” IEEE Conference on Decision and Control (CDC), pp. 3971–3976, Osaka, Japan, December 15–18, 2015. [pdf] [doi] [slides]
• M. B. Horowitz, I. Papusha, and J. W. Burdick “Domain Decomposition for Stochastic Optimal Control,” IEEE Conference on Decision and Control (CDC), pp. 1866–1873, Los Angeles, CA, December 15–17, 2014. [pdf] [arXiv] [doi]
• I. Papusha, E. Lavretsky, and R. M. Murray. “Collaborative System Identification via Parameter Consensus,” American Control Conference (ACC), pp. 13–19, Portland, OR, June 4–6, 2014. [pdf] [doi] [slides]

Earlier Work

Reports

• I. Papusha. “Fast Automatic Background Extraction via Robust PCA,” 2011. [pdf] [pcp_admm.m code]
• I. Papusha and M. Ho. “Hough Transform for Directional Orientation,” 2010. [pdf]

Talks

• “Affine Multiplexing Networks,” Applications of Formal Methods to Control Theory and Dynamical Systems (FoMA Workshop), Carnegie Mellon University, June 23, 2018. [pdf]
• “Automata Theory and Dynamic Programming,” SIAM Conference on Applications of Dynamical Systems (DS17), Snowbird, UT, May 21, 2017.
• “Toward Learning and Adaptation in Optimization Based Control,” Google, Mountain View, CA, January 12, 2016.
• “Networked Adaptive Systems,” GRASP Lab, University of Pennsylvania, January 17, 2014.