Tools
Sorted by:
Try your query at:
 |
 |
 |
 |
 |
 |
Results 1 - 10
of
108,129
Table 1 Mean differences in proportion by industry. Cells show the difference (I-J) for proportion of matched financial statement terms, with significant differences bolded. For example, mean proportion of matched line times for Beverages minus mean proportion matched for Airlines = 9.379 percent, and is significant. Airlines significantly differs from many industries, suggesting a candidate for an industry-specific XBRL taxonomy extension.
"... In PAGE 19: ... For example, financial statements from 80 companies in 12 industries, parsed with FRAANK, show proportions of matched line items by industry that range from 54% to 70% (Table 3), and by-statement matching proportions of approximately 64% (Table 4). Factorial analysis of these results shows that the percentage of items matched for the Airlines industry is significantly different from 7 out of 11 of the remaining industries, suggesting a possible candidate for an industry-specific XBRL-taxonomy extension ( Table1 ), as well as a possible need for additional synonyms in the knowledge-base. As discussed elsewhere (Bovee et al.... ..."
Table 5. Mean values and deviations of quantization error of different algorithms. Most left column indicates how many clusters were formed. Abbreviations: mn = mean value, dv = deviation, HKM = hard k-means clustering algorithm, FKM = fuzzy k-means clustering algorithm, SOM = Kohonen self-organizing feature map.
TABLE 7 IMPORTANCE OF PERSONAL FACTORS TO HIRING DECISION
Table 5. Mean value of many indicators in the second part of the experiment by treatment treatment 1 treatment 2 treatment 3 P- Value F test ANOVA p-value Kruskal-
"... In PAGE 9: ... In this paper, instead, the attention is focused on the satisfaction and on its relation with score10. Table5 compares many mean values11 of the three groups. Among the treatments there are no significant differences in mean total score.... ..."
Table 14: Summary of performance differences of all experiments. The first number indicates the number of times the model of the row had a higher mean performance than the model of the column. The second number indicates how many times this higher performance was statistically significant in itself.
2006
"... In PAGE 22: ...Analysis In this section, we perform an analysis of the most important experimental results. Table14 is a summary of the results of all fourteen experiments con- cerning the performance differences between the models (Section 6). The first number shows for each model in how many experiments it had a higher mean performance than each other model.... ..."
Table 1. Probes for yeast cdc28 gene generated by ProbeSelect*
2000
"... In PAGE 5: ....2. Probes generated by ProbeSelect Probes are generated for each gene by ProbeSelect. Ten probes for the yeast CDC28 gene are shown in Table1 . Ta- ble 2 describes the detailed information about each probe for this gene and a part of match sequences with 4 or fewer mismatches.... In PAGE 5: ... In table 2, a0a103a1 and a3a5a4 for some probe hybridiza- tion structures of yeast gene CDC28 are listed. Of the ten probes listed in Table1 , only probe-6 has an alternative tar- get in the genome with a free energy difference within 10 kcal/mole of the true target free energy. Based on the data in Table 2 one could pick the optimum probe (or small set of probes) for the CDC28 gene.... ..."
Cited by 5
Table 1 shows the mean results for each source density on each row, with N = 250, the number of input points, and 100 replications of each experiment. The best performing algorithm on each row is shown in bold face. Note that RADICAL performs best in 10 of 18 experiments, substantially outperforming the second best in many cases. The mean performance in these experiments is shown in the row labeled mean, where RADICAL has lower error than all other algorithms tested. The final row of the table represents experiments in which two (generally different) source densities were chosen randomly from
2003
"... In PAGE 18: ...0 5.8 Table1 : The Amari errors (multiplied by 100) for two-component ICA with 250 samples. For each pdf (from a to r), averages over 100 replicates are presented.... ..."
Cited by 4
Table 1 shows the descriptive statistics for the naturalness of concepts. We use the following statistical measurements for our analysis: Mean, the mean value of the number of search results for a concept; Standard Deviation; Range, the difference between the minimum and the maximum. The ontology with the highest mean value, 5,059,901 is the Semantic Network and the ontology with the lowest mean value, 21,499 is the Metathesaurus of the UMLS. This agrees with our intuition, as the Metathesaurus contains many highly specialized medical terms. Note that the sample sizes are quite different. For large ontologies/terminologies, sample sizes of about 3000 were used, because processing the whole terminology was beyond our computational resources. However, we were able to process the complete UMLS Semantic Network, as it is small. Also, the part of the Metathesaurus consisting of all concepts with IS-A relationships was processed.
"... In PAGE 6: ... Table1 : The descriptive statistics Figure 4 shows information about the frequency of single-word labels and multi-word labels for each of our ontologies.... ..."
Table 5. Multiple comparison of Tukey HSD for Keyword accuracy
"... In PAGE 19: ... Because of this, post hoc multiple comparisons (Tukey honestly significant differences (HSD)) were conducted to evaluate pair-wise differences among the means. The data displayed in Table5... ..."
Results 1 - 10
of
108,129
