Keywords: numerical linear algebra, bootstrap, randomized algorithms, sketching
Randomized Numerical Linear Algebra (RandNLA) is an interdisciplinary research area that exploits randomization as a computational resource to develop improved algorithms for large-scale linear algebra problems. While the motivating applications for RandNLA are in large-scale machine learning and data analysis, most work in RandNLA so far has come from the perspectives of theoretical computer science and numerical linear algebra. This has begun to change, and many of the most exciting current developments in RandNLA have to do with focusing on statistical and optimization considerations. Here, we describe recent results that use the bootstrap to enhance error estimation in the contexts of large-scale matrix multiplication and least squares.