Algorithms for Semantic Meta-Mining
|Director of thesis||Prof. Christian Pellegrini|
|Co-director of thesis||Melanie Hilario, Alexandros Kalousis|
|Summary of thesis||
We focus on a new meta-learning framework where in addition to dataset descriptors we have data mining algorithms and workflows descriptions. We develop new similarity measures between datasets and algorithms or workflows supervised by their relative performance and we demonstrate the applicability of our framework on planning to learn problems.
|Administrative delay for the defence|