Memory-Efficient Modewise Measurements for Tensor Compression and Recovery

Explore a comprehensive lecture on memory-efficient modewise measurements for tensor compression and recovery presented by Liza Rebrova from Princeton University at IPAM's Tensor Networks Workshop. Delve into the importance of data-oblivious measurements in low-rank data compression and recovery techniques, particularly in streaming settings and iterative algorithms. Examine the challenges of creating sketches that reflect tensor structure while maintaining efficiency. Discover recent developments in flexible and provable modewise sketches for tensor data processing, including compressed CP rank fitting, modewise tensor iterative hard thresholding, and direct recovery from leave-one-out modewise measurements for low Tucker rank tensors. Gain valuable insights into advanced tensor analysis techniques and their applications in data compression and recovery.

Liza Rebrova - Memory-efficient modewise measurements for tensor compression and recovery