Seminar: Coverage Algorithms in Streams

There are numerous settings in which we need to deploy servers, warehouses, or facilities that maximise the number of clients they serve. It turns out that a fairly clean algorithm is the “right” approach for this problem. However, the data about which client would be satisfied by which facility might not be received all at once, but in a stream, or sequence, in which we only have capacity to remember some of the data. Realizing this clean algorithm in this computational model requires some innovation. I recap several of my works on so-called coverage problems in streams, and hope to portray some of the key insights, tools, and computational thinking behind streaming methods. Time permitting, I hope this opens up some discussion about computational modelling of memory. This talk includes research that was collaborations with Graham Cormode, Howard Karloff, Amit Chakrabarti, Stephen Jaud, Farhana Choudhury, Rowan Warneke, Phil Cervenjak, Junhao Gan, and William Umboh.

Tony Wirth is Professor and Deputy Head in the School of Computer Science at The University of Sydney. His PhD was at Princeton University, working on approximation algorithms for correlation clustering.Immediately following, he returned to his alma mater, The University of Melbourne, for a 19-year career as an academic, working on algorithms for binary search with errors, text archiving, streaming problem, clustering, and covering.

Alternatively, join online:

Microsoft Teams
Meeting ID: 347 752 625 007 6
Passcode: wP7PG2uh