Small Ball Probabilities for Random Tensors and Analysis of Tensor Decompositions

Explore a 53-minute conference talk on small ball probabilities for simple random tensors and their applications to smoothed analysis of tensor decompositions. Delve into sharp anticancellation inequalities for simple random tensors under minimal randomness assumptions. Discover how these findings are applied to revisit known algorithms for reconstructing tensors using simple ones. Presented by Grigoris Paouris from Texas A&M University, College Station, this talk is part of IPAM's Tensor Networks Workshop. Gain insights into joint work with Xuehan Hu and expand your understanding of tensor analysis and decomposition techniques.