Towards User-Centered Active Learning Algorithms

dc.contributor.authorBernard, Jürgenen_US
dc.contributor.authorZeppelzauer, Matthiasen_US
dc.contributor.authorLehmann, Markusen_US
dc.contributor.authorMüller, Martinen_US
dc.contributor.authorSedlmair, Michaelen_US
dc.contributor.editorJeffrey Heer and Heike Leitte and Timo Ropinskien_US
dc.date.accessioned2018-06-02T18:07:26Z
dc.date.available2018-06-02T18:07:26Z
dc.date.issued2018
dc.description.abstractThe labeling of data sets is a time-consuming task, which is, however, an important prerequisite for machine learning and visual analytics. Visual-interactive labeling (VIAL) provides users an active role in the process of labeling, with the goal to combine the potentials of humans and machines to make labeling more efficient. Recent experiments showed that users apply different strategies when selecting instances for labeling with visual-interactive interfaces. In this paper, we contribute a systematic quantitative analysis of such user strategies. We identify computational building blocks of user strategies, formalize them, and investigate their potentials for different machine learning tasks in systematic experiments. The core insights of our experiments are as follows. First, we identified that particular user strategies can be used to considerably mitigate the bootstrap (cold start) problem in early labeling phases. Second, we observed that they have the potential to outperform existing active learning strategies in later phases. Third, we analyzed the identified core building blocks, which can serve as the basis for novel selection strategies. Overall, we observed that data-based user strategies (clusters, dense areas) work considerably well in early phases, while model-based user strategies (e.g., class separation) perform better during later phases. The insights gained from this work can be applied to develop novel active learning approaches as well as to better guide users in visual interactive labeling.en_US
dc.description.number3
dc.description.sectionheadersHigh-dimensional Data
dc.description.seriesinformationComputer Graphics Forum
dc.description.volume37
dc.identifier.doi10.1111/cgf.13406
dc.identifier.issn1467-8659
dc.identifier.pages121-132
dc.identifier.urihttps://diglib.eg.org:443/handle/10.1111/cgf13406
dc.identifier.urihttps://doi.org/10.1111/cgf.13406
dc.publisherThe Eurographics Association and John Wiley & Sons Ltd.en_US
dc.subjectI.3.3 [Computer Graphics]
dc.subjectPicture/Image Generation
dc.subjectLine and curve generation
dc.titleTowards User-Centered Active Learning Algorithmsen_US
Files
Collections