Results 11 - 20 of 279,204

Table IV.1: Training set accuracy only for the newly added task. Figures in parentheses denote standard error of the mean. Expert task Greebles training set accuracy(%) Non-expert 71.2(2.00)

in Visual Expertise Is A General Skill
by Maki Sugimoto, Maki Sugimoto, All Rights Reserved 2001
Cited by 3

Table 3.2: Non-expert human players trained Sophie with and without guidance communica- tion available and also show positive effects of guidance on the learning performance. (F = failed trials, G = first success). The following are the results of 1-tailed t-tests.

in Socially guided machine learning
by Andrea Lockerd Thomaz 2006
Cited by 8

Table 8. Agreement proportions (%) based on 10 non-experts.

in Focal Accent -- f_0 Movements and Beyond
by Heldner Mattias 2001
"... In PAGE 49: ....3.2. Agreement proportions; comparisons with ToBI evaluations Table8... ..."

Table 8. Agreement proportions (%) based on 10 non-experts.

in Labelling of Boundaries and Prominences By Phonetically Experienced and Non-Experienced Transcribers
by Eva Strangert, Mattias Heldner
"... In PAGE 18: ...3.2 Agreement proportions; comparisons with TOBI evaluations Table8 shows agreement proportions calculated as for the experts (see 3.... ..."

Table 1: User ratings concerning general topics.

in Evaluating Multi-modal Input Modes in a Wizard-of-Oz Study for the Domain of Web Search
by Ra Klein, Ingrid Schwank, Michel Généreux, Harald Trost, Alexandra Klein 2001
"... In PAGE 7: ... Generally, expert and non-expert users were faster using all three input modes (cf. Table1 and Table 2). So far, results indicate that multi-modal interaction can provide a usable and efficient access to documents on the Web, especially for non-experts.... ..."
Cited by 3

Table 4: expert-net ensemble performance as diverse multiversion systems The data in table 4 is derived from the same LIC1 expert-net ensemble as the previous results (table 1), except this time the individual experts were tested on the same 1000 patterns randomly selected from the complete function (i.e. they were treated as non- experts). Three di erent sets of 1000 test patterns gives the three rows of results. A comparison between these two sets of results reveals that the expert networks per- formed on average 20% worse when treated as non-experts. This result suggests that the 9

in Self-Organizing Sets of Experts
by Niall Griffith, Derek Partridge

Table 3: Error measurements for segmentation of clinical MRI cases.

in multi-phase three-dimensional implicit deformable
by Elsa D. Angelini A, Ting Song A, Brett D. Mensh B, Andrew Laine A
"... In PAGE 10: ... The segmentation method was able to handle multiple challenges without any a priori information or shape constraints that include the extraction of highly-convoluted white matter surfaces, the extraction of separate ventricular structures for the CSF, and handling of different volume sizes of the three structures in a simultaneous segmentation scheme. Error measurements for the segmentation of the three clinical cases are reported in Table3 . These results report overall clinically satisfactory (useful) performance of the proposed segmentation method.... ..."

Table 9a. Inter-rater Agreement T (Identical judgements) based on 10 non-experts.

in Focal Accent -- f_0 Movements and Beyond
by Heldner Mattias 2001
"... In PAGE 50: ... Table9 b. Inter-rater Agreement T (One level differences) based on 10 non-experts.... ..."

Table 9a. Inter-rater Agreement T (Identical judgements) based on 10 non-experts.

in Labelling of Boundaries and Prominences By Phonetically Experienced and Non-Experienced Transcribers
by Eva Strangert, Mattias Heldner
"... In PAGE 18: ... Table9 b. Inter-rater Agreement T (One level differences) based on 10 non-experts.... ..."

(Table 2). We distinguish topics by a clear shift into a different topic, as interpreted by a non-expert.

in Generating Reports from Case-Based Knowledge Artifacts’,in
by Rosina O. Weber, Sidath Gunawardena, Jason M. Proctor 2007
Cited by 1
Results 11 - 20 of 279,204