Postdoctoral Fellow

Department of Mathematics

The University of British Columbia

Leonard S. Klinck 311d The University of British Columbia 2329 West Mall Vancouver, BC Canada V6T 1Z4
b.joshi@math.ubc.ca My Google scholar page
I am a postdoctoral fellow in the Department of Mathematics at the University of British Columbia. I am very fortunate to be mentored by Professors Yaniv Plan, Ozgur Yilmaz, and Micheal Friedlander. I have worked on signal recovery problems in clustering and parameter estimation. I use ideas and results from high dimensional geometry and high dimensional probability to find and analyze convex and non-convex programs; providing a theoretical recovery guarantee for these programs. Implementations of some of these algorithms in Python/Julia can be found in GitHub.

Publications

PLUGIn: A simple algorithm for inverting generative models with recovery guarantees, Advances in Neural Information Processing Systems, 2021. link (Spotlight Presentation, as awarded to top 3% of submissions.)
PLUGIn-CS: A simple algorithm for compressive sensing with generative prior, NeurIPS 2021 Workshop Deep Inverse Program, 2021. link
Polar deconvolution of mixed signals, 2020. (In review) arXiv
Bilinear compressed sensing under known signs via convex programming, 2019. arXiv
Global guarantees for blind demodulation under generative prior, Advances in Neural Information Processing Systems, 2019. arXiv
A convex program for bilinear inversion of sparse vectors. Advances in Neural Information Processing Systems, 2018. arXiv
BranchHull: Convex bilinear inversion from the entrywise product of signals with known signs. Applied and Computational Harmonic Analysis, 2018. arXiv
A convex program for mixed linear regression for well-separated data. Information and Inference: A Journal of IMA, 2017. arXiv

Teaching

Term
Course Webpage
Institution
2021W Term 2
Integral Calculus (Math 101)
UBC
2020W Term 1
Differential Calculus (Math 104)
UBC
2020S Term 2
UBC
2018 Fall
Matrix Analysis (CAAM 335)
Rice