Search results for "Data description"
showing 4 items of 4 documents
Discovering single classes in remote sensing images with active learning
2012
When dealing with supervised target detection, the acquisition of labeled samples is one of the most critical phases: the samples must be yet representative of the class of interest, but must also be found among a vast majority of non-target examples. Moreover, the efficiency of the search is also an issue, since the samples labeled as background are not used by target detectors such as the support vector data description (SVDD). In this work we propose a competitive and effective approach to identify the most relevant training samples for one-class classification based on the use of an active learning strategy. The SVDD classifier is first trained with insufficient target examples. It is t…
Reproducibility of optical biometry using partial coherence interferometry: intraobserver and interobserver reliability
2001
Abstract Purpose: To evaluate the intraobserver and interobserver variability in axial length (AL), anterior chamber depth (ACD), and corneal radius measurements using an optical biometry instrument based on partial coherence interferometry (PCI). Setting: Johannes Gutenberg University, Mainz, Germany. Methods: In this observational case series and interobserver reliability trial, 30 healthy, emmetropic to moderately myopic eyes of 15 volunteers were evaluated. The AL, ACD, and corneal radius were measured 20 times in 10 eyes by 1 observer to evaluate the intraobserver variability. To evaluate the interobserver variability, the measurements were taken in 20 eyes by 5 different observers. Me…
I: Boxplots - die flexible Alternative zum „Antennenbildchen”
2002
Data description for continuous parameters is sometimes only based on means and standard deviations of measurement series, graphical representation only concentrates on corresponding "MSE plots", which provide means and standard deviations or even only mean squared errors. However, this strategy is only correct for normally distributed data. Outliers may seriously bias mean and standard deviations and may therefore lead to wrong clinical conclusions. The present paper suggests the use of medians and quartiles, which - just like their graphical pendant, the nonparametric "box whisker plot" - can be applied much more flexibly.
Adaptive semantic web based environment for web resources
2008
Tulevaisuuden kaikkialle ulottuvassa internetissä tietojärjestelmät kommunikoivat paitsi käyttäjien kanssa, myös toisten sovellusten ja instrumentoitujen laitteiden kanssa. Tämän dynaamisen ja heterogeenisen digitaalisen ympäristön hallinnoimiseksi ja hyödyntämiseksi käyttämiemme laitteiden tulisi olla nykyistä proaktiivisempia ja tiedon tulisi olla kuvattu nykyistä kontekstitietoisemmalla tavalla.Lisäksi tulevaisuuden verkon resurssit tulisi kyetä kuvaamaan semanttisesti, jotta eri resurssit voidaan löytää ja sovittaa yhteen automaattisesti. Tämä mahdollistaa myös päätelmien tekemisen järjestelmien tiedoista sekä monimutkaisen kokonaisuuden komponenttien käyttäytymisen ohjaamisen helpommin…