F-Measure Optimisation in Multi-label Classifiers

TitleF-Measure Optimisation in Multi-label Classifiers
Publication TypeConference Paper
Year of Publication2012
AuthorsPillai, I, Fumera, G, Roli, F
Conference Name21st International Conference on Pattern Recognition
Date Published11-15/11/2012
Conference LocationTsukuba, Japan
Keywordsdoc00, doc01, F-measure, multi-label categorization, thresholding

When a multi-label classifier outputs a real-valued score for each class, a well known design strategy consists of tuning the corresponding decision thresholds by optimising the performance measure of interest on validation data.
In this paper we focus on the F-measure, which is widely used in multi-label problems.
We derive two properties of the micro-averaged F measure, viewed as a function of the threshold values, which allow its global maximum to be found by an optimisation strategy with an upper bound on computational complexity of O(n2 N2), where N and n are respectively the number of classes and of validation samples.
So far, only a suboptimal threshold selection rule and a greedy algorithm without any optimality guarantee were known for this task.
We then devise a possible optimisation algorithm based on our strategy, and evaluate it on three benchmark, multi-label data sets.

Citation Key 1349
pillai_ICPR2012_with_proofs.pdf295.91 KB