Home // ADVCOMP 2020, The Fourteenth International Conference on Advanced Engineering Computing and Applications in Sciences // View article
Authors:
Tim vor der Brück
Keywords: F-Measure;paradox; precision; recall; NaN
Abstract:
Abstract—Paradoxes have raised a lot of interest in mathematics and computer science. What fascinates people about it is that such a paradox contains a self-contradictory statement that dissents with usual believes and expectations. The range of discovered paradoxes is long. One of the most famous is probably the proposition of Russell that states that no set can exist that contains all sets that do not contain itself as a subset. The paradox arises in the proof, where it is shown that such a set must contain itself if and only if it does not contain itself. In this paper, we derive a paradox about the F-measure, one of the most important metrics in machine learning. The contribution of this paper is two-fold. On the one hand, we investigate typical properties of the FMeasure, on the other hand, we show that they are contradictory and therefore constitute a paradox, to several properties of the harmonic mean, where the F-Measure is a special case of.
Pages: 1 to 5
Copyright: Copyright (c) IARIA, 2020
Publication date: October 25, 2020
Published in: conference
ISSN: 2308-4499
ISBN: 978-1-61208-812-9
Location: Nice, France
Dates: from October 25, 2020 to October 29, 2020