Speaker
Dr
Sebastian Buschjäger
(Lamarr Institute, TU Dortmund)
Description
Submodular Function Maximization arises in many different applications fields in Machine Learning and Data Science, especially in Data Summarization. In Data Summarization, we want to select a meaningful subset of the data points for human inspection so the user can get a quick overview of meaningful data points. We contributed a novel Submodular Function Maximization algorithm called ThreeSieves that can extract data summaries on the fly while the data is generated, allowing humans to inspect data in near real time.
