Ocean Engineering and Applied Technology

Oscar López, Ph.D.
Assistant Research Professor
772.242.2410
lopezo@skyupiradio.com
Publications
ORCiD profile

Research interests include:

• Information theory and signal processing
• Tensor-based statistical inference, high-dimensional data analysis and processing
• Applications to oceanographic and atmospheric data; compression, de-noising and processing
• Computational imaging via compressive sensing techniques
• Hyperspectral imaging
• Seismic data acquisition and imaging
• Underwater imaging in turbid conditions; single-photon counting and lidar

Dr. López's current focus is in multidimensional data analysis and processing. The main goal is to exploit the high dimensional structure of datasets in order to improve the efficiency of inference methods. Achieving this task is key to combat the “curse of dimensionality” due to the deluge of information emanating from evolving data collection technologies. This research views this curse as a blessing, where the “concentration of measure” phenomena in high dimensions reveals the pervasive simplicity of multi-dimensional volumes and provides a promising avenue to future-proof data analytics.

In particular, this research seeks to:

1. Embrace the multidimensional nature of data. Demonstrate and quantify the advantages of tensor-based processing when compared to lower dimensional approaches, e.g., that consider vector/matrix reshapings or slices of arrays.
2. Study sample complexities of low-rank tensor recovery problems. Establish optimal rates for tensor completion methods that apply the canonical polyadic decomposition and other factorization choices.
3. Demonstrate the compression capabilities of tensor decompositions when applied to oceanographic and atmospheric data. Develop decomposition-contrained optimization programs for computational imaging and related inverse problems.

Research opportunities are available at the graduate and postdoctoral level, contact Dr. López with your CV for more information.

Featured Publications and Preprints

O. López, R. Lehoucq, C. Llosa-Vite, A. Prasadan, D. Dunlavy. 2024. "The Average Spectrum Norm and Near-Optimal Tensor Completion". http://arxiv.org/abs/2404.10085

O. López, A. Ernce, B. Ouyang, E. Malkiel, C. Gong, M. Twardowski. 2024. "Advancements in Compressive Hyperspectral Imaging: Adaptive Sampling with Low-Rank Tensor Image Reconstruction". http://www.researchgate.net/publication/379832376_Advancements_in_Compressive_Hyperspectral_Imaging_Adaptive_Sampling_with_Low-Rank_Tensor_Image_Reconstruction

O. López, R. Kumar, N. Moldoveanu and F. J. Herrmann, "Spectral Gap-Based Seismic Survey Design," in IEEE Transactions on Geoscience and Remote Sensing, vol. 61, pp. 1-9, 2023, Art no. 5901809, http://ieeexplore.ieee.org/document/10018252

K. D. Harris, O. López, A. Read and Y. Zhu, "Spectral gap-based deterministic tensor completion," 2023 International Conference on Sampling Theory and Applications (SampTA), New Haven, CT, USA, 2023, pp. 1-6, http://ieeexplore.ieee.org/document/10301415

O. López, D. Dunlavy, R. Lehoucq. "Zero-Truncated Poisson Regression for Sparse Multiway Cound Data Corrupted by False Zeros". Information and Inference: A Journal of the IMA, Volume 12, Issue 3, September 2023, Pages 1573–1611, http://doi.org/10.1093/imaiai/iaad016

O. López. "Near-Optimal Weighted Matrix Completion". Journal of Machine Learning Research, 24(283), 2023, Pages 1-40, http://jmlr.org/papers/v24/22-0331.html