### Table 2: Categorization of the error regions and errors found in the NAB test data.

1997

"... In PAGE 4: ...) (c) If neither of the above work out, then we categorize the error as belonging to class Miscellaneous. Table2 shows the distribution of categorizations for the NAB de- velopment data, a total of 1523 utterances. 5.... ..."

Cited by 8

### Table 4: Absolute and relative errors in the final 16-region segmentation map. Region Size in the Errors In %

"... In PAGE 58: ...82% (32239 pixels with different labeling among 262144 pixels in the lattice). Table4 gives absolute and relative errors for each region in the test image. It is easily seen that most differences between the ideal and segmentation maps are due to similarities between subparts of different weakly homoge- neous textures.... ..."

### Table 1. Discretization error for arctanh regions.

"... In PAGE 10: ... Here the conformal map is given by f(z) = log((1+rz)=(1?rz)); 0 lt; r lt; 1; which maps the disk to increasingly elongated, cigar-shaped regions as r quot; 1: This map is perhaps the simplest example of a conformal map exhibiting the exponential crowding [De]. Table1 shows that the discrete error behaves like O(Rn) as expected. The semilog plot in Figure 1b exhibits the superlinear convergence behavior of the residuals in the conjugate gradient iterations as an illustration of Theorem 3(i).... In PAGE 10: ... We have also computed kxq+1 ? xqkAn as an approximation of keqkAn for our examples and found the values to behave roughly like the residuals. Also note that the number of outlying eigenvalues of In + Mn in Figure 1a is roughly equal to the number of conjugate gradient iterations given in Table1 and Figure 1b, as one would expect. Similar results were obtained for this and other examples with analytic boundaries in [CDH] by solving the normal equations.... ..."

Cited by 1

### Table 9. Regionally averaged trends and standard errors for flve regions of USA, and overall.

2003

"... In PAGE 38: ...same regions in a slightly difierent context) shows the 187 stations divided into flve regions. In Table9 , we give estimated \regional trends quot;, with standard errors. The regional trends are derived by averaging the smoothed fl1 estimates over a very dense grid of points within the region of interest, and the standard errors are the square roots of the estimated variances of those averages, given the ^ fl1 values.... ..."

### Table 4.14 dF/dx s.e. dF/dx s.e. dF/dx s.e. dF/dx s.e.

in Vulnerability in Consumption, Education and Health: Evidence From Moldova During the Russian Crisis

Cited by 1

### Table 4.4: Classification of error regions according to [21].

### Table 4.5: Classification of error regions in MFCC and AF systems.

### Table 13. Regionally averaged trends and standard errors for five regions of USA, and overall.

2005

Cited by 2