(she/her/hers)
Northwestern University
Algorithms, beyond-worst-case, scheduling
Sami is a post-doctoral researcher in Theoretical Computer Science at Northwestern University, and before that she obtained her PhD in 2021 from UW. Sami primarily works remotely from Seattle. She is a member of the 2021 class of CIFellows. Broadly speaking, Sami like problems in combinatorial optimization, think hypergraph matchings, coloring, and scheduling stuff. Her post-doc work has focused on beyond-worst-case analysis.
Beyond-worst-case optimization
Often, theoretical worst-case performance bounds (e.g. approximation and competitive ratios, upper bounds on runtimes and sample complexities, etc.) are far worse than what one can achieve in practice. One reason this happens is that in practice instances have certain observable, underlying structure that an algorithm can exploit. For example, in an online scheduling problem, the jobs that arrive today might look very similar to the jobs that arrived yesterday. The exploding field of learning-augmented algorithms is concerned with designing algorithms that circumvent worst-case bounds by using learned predictions about the input.