Summarizing Massive Data Sets via Large-Scale Submodular Maximization - Prof. Andreas Krause

Yandex School of Data Analysis Conference Machine Learning: Prospects and Applications How can one summarize a massive data set without being able to fit it in the main memory of a single computer? I will present recent work on extracting representative elements from a large set of data. In particular, I will discuss selecting a subset of, say, k points from data arriving as a stream or spread across a cluster, which are most representative according to some objecti