Research

Papers on optimal transport, sampling, optimization, and machine learning. See also Google Scholar.

  1. Simplifying Optimal Transport through Schatten-p Regularization

    Maunu, T. · Transactions on Machine Learning Research

    Paper arXiv

  2. Subspace Langevin Monte Carlo

    Maunu, T. and Yao, J. · SIAM Journal on Mathematics of Data Science 8(3):850–886

    Paper arXiv

  3. A Smoothing Newton Method for Rank-One Matrix Recovery

    Maunu, T. and Abreu, G. · Under review

    arXiv

  4. Global Convergence of Iteratively Reweighted Least Squares for Robust Subspace Recovery

    Lerman, G., Li, K., Maunu, T., and Zhang, T. · Under review

    arXiv

  5. Acceleration and Implicit Regularization in Gaussian Phase Retrieval

    Maunu, T. and Molina-Fructuoso, M. · Proceedings of the 27th International Conference on Artificial Intelligence and Statistics, PMLR 238:4060–4068

    Paper arXiv

  6. First-Order Algorithms for Optimization over Graph Laplacians

    Maunu, T. · Proceedings of the 14th International Conference on Sampling Theory and Applications (SampTA), IEEE

  7. Bures-Wasserstein Barycenters and Low-Rank Matrix Recovery

    Maunu, T., Le Gouic, T., and Rigollet, P. · Proceedings of the 26th International Conference on Artificial Intelligence and Statistics, PMLR 206:8183–8210

    Paper arXiv

  8. Depth Descent Synchronization in SO(d)

    Maunu, T. and Lerman, G. · International Journal of Computer Vision 131(4):968–986

    arXiv

  9. Stochastic and Private Nonconvex Outlier-Robust PCA

    Maunu, T., Yu, C., and Lerman, G. · Mathematical and Scientific Machine Learning, PMLR, pp. 173–188

    arXiv

  10. Scalable Cluster-Consistency Statistics for Robust Multi-Object Matching

    Shi, Y., Li, S., Maunu, T., and Lerman, G. · Proceedings of the 2021 International Conference on 3D Vision (3DV), IEEE, pp. 352–360

    arXiv

  11. Score-Based Generative Neural Networks for Large-Scale Optimal Transport

    Daniels, M., Maunu, T., and Hand, P. · Advances in Neural Information Processing Systems 34:12955–12965

    arXiv

  12. SVGD as a Kernelized Wasserstein Gradient Flow of the Chi-Squared Divergence

    Chewi, S., Le Gouic, T., Lu, C., Maunu, T., and Rigollet, P. · Advances in Neural Information Processing Systems 33:2098–2109

    Paper arXiv Code

  13. Exponential Ergodicity of Mirror-Langevin Diffusions

    Chewi, S., Le Gouic, T., Lu, C., Maunu, T., Rigollet, P., and Stromme, A. · Advances in Neural Information Processing Systems 33:19573–19585

    arXiv

  14. Gradient Descent Algorithms for Bures-Wasserstein Barycenters

    Chewi, S., Maunu, T., Rigollet, P., and Stromme, A. J. · Proceedings of the Conference on Learning Theory, PMLR, pp. 1276–1304

    arXiv

  15. Robust Subspace Recovery with Adversarial Outliers

    Maunu, T. and Lerman, G. · arXiv preprint arXiv:1904.03275

    arXiv

  16. A Well-Tempered Landscape for Non-Convex Robust Subspace Recovery

    Maunu, T., Zhang, T., and Lerman, G. · Journal of Machine Learning Research 20(37)

    arXiv

  17. An Overview of Robust Subspace Recovery

    Lerman, G. and Maunu, T. · Proceedings of the IEEE 106(8):1380–1410

    arXiv

  18. Fast, Robust and Non-Convex Subspace Recovery

    Lerman, G. and Maunu, T. · Information and Inference: A Journal of the IMA 7(2):277–336

    arXiv