Explore a 53-minute lecture on sketching and projecting methods for linear systems and linear discriminant analysis. Delve into recent work presented by Deanna Needell from the University of California, Los Angeles. Discover how stochastic gradient approaches are utilized to obtain solutions similar or identical to un-sketched problems while significantly reducing computational burden. Examine convergence guarantees for sketched predictions on data within a fixed number of iterations, taking into account both modeling assumptions and algorithmic randomness from the sketching procedure. Gain insights into numerical results comparing these approaches and their applications in algorithm design.