hide
Free keywords:
-
Abstract:
The life sciences create new challenges for feature selection in data mining: First, there is a need for feature selection on structured
data such as strings and graphs. Second, a deeper theoretical understanding of the connections between existing feature selection approaches would be beneficial, to explain the discrepancies in their results on the same datasets. Third, the large number of features poses a computational and algorithmic challenge and requires the
development of new, efficient selection techniques. In this talk, we will present our work on these three topics.
