### Table 9: Project Evaluations by Percentage Abstract accurate and 1 2 3 4 5

1991

Cited by 1

### Table 4 Comparison of exchange and correlation energies of Be for LDA and the various GGA apos;s with the corresponding accurate values. The approximate functionals are evaluated for the exact (not the self-consistent) densities. The numbers in parentheses are the percentage errors. Energies are in Hartree atomic units. Taken from Ref. [47]. Functional

### Table 5 Comparison of LDA, Perdew-Wang apos;91 GGA and accurate values for various components of the Kohn-Sham total energy, the density-weighted integral of vxc and single-particle eigenvalues of Be. The approximate functionals are evaluated for the respective self- consistent densities. The accurate values probably have errors in the last digit quoted. The accurate value of the total energy is from Ref. [70]. Energies are in Hartree atomic units. Taken from Ref. [47].

"... In PAGE 22: ... It is di cult for any GGA functional to mimic the correct Z dependence of both iso-electronic series.In Table5 , we show various quantities, computed by LDA and the Perdew-Wang apos;91 GGA, and the corresponding accurate values for Be. LDA and GGA values were obtained through a self-consistent procedure and are therefore evaluated for the LDA and the GGA density respectively.... In PAGE 22: ... The change in Ex and Ec due to self-consistency are not negligible. The values of Ex and Ec in Table5 di er from those in Table 5 by a larger amount for LDA than for the Perdew-Wang apos;91 GGA re ecting our earlier observation that the GGA yields somewhat more accurate self-consistent densities than does LDA. The last two columns show the errors in LDA and GGA.... ..."

### Table 8 Results for a single level Optimization using the accurate bump.

1998

"... In PAGE 8: ... The comparative improvement of each method using any of the three proposed strategies may then be calculated, see Table 9. Here the results are normalised by diving the average performance of each method in Tables 2 to 7 by the values in Table8 . The results may also be normalised by dividing by the best ever results obtained for the accurate function evaluation (0.... ..."

Cited by 2

### Table 4. Evaluation of Attachment Resolution

"... In PAGE 7: ... This more accurate evaluation gives us more significant information on the system capabilities concerning attachment resolution. Results Table4 reports the test scores concerning the precision and recall of the two comparative experiments performed. The total precision of our method reaches more than 90%, whereas the total recall is about 75%.... ..."

### Table 2: Experimental Parameters number of threads. Positive numbers indicate the simulator overestimates the actual system. On average, the simulator is 5.1% slower than the actual system and the difference ranges from 3.9% faster to 27.2% slower. The difference between the actual system and the simulator tends to decrease as the number of threads increases and the query arrival rate decreases because we slightly overestimate the query evaluation time. In general, the simulator matches the actual system closely, but we are not able to accurately model every query. Due to our simple model it is difficult to reduce the amount of variation between our simulator and the actual system times. Given that our simulator evaluates most queries accurately, especially for 4 threads (average 1.2% error), we believe it is not worth the added complexity to change the model.

2000

"... In PAGE 16: ...3 Experimental Parameters In this section, we describe the parameters we use in our simulation experiments. Table2 presents the parameters, their values, and abbreviations. We describe each parameter in more detail below.... In PAGE 16: ... We model the command rate using a Poisson distribution. Table2 lists the different distribution averages that we use in our experiments. The slowest rate averages one command every ten seconds and the fastest rate averages ten commands per second.... In PAGE 17: ...up to 128 Inquery servers (128 GB of data). Terms Per Query (TPQ) Table2 shows the three different average query lengths we use in our experi- ments. We obtain two of the values from the 103rd Congressional Record (2 terms) and TIPSTER 1 query set (27 terms).... ..."

Cited by 32

### Table 1: Averaged Error of Models with respect to accurate simulation

"... In PAGE 4: ... To evaluate the ac- curacy of the widely used WATTCH (3) model, the values from WATTCH were scaled to 130nm and 90nm technol- ogy. Table1 shows the error margin of the proposed em- pirical model and the WATTCH model with respect to the simulation of the actual designs. The error values were ob- tained for the complete design range and are averaged out.... ..."

### Table 7 Evaluation Results

"... In PAGE 6: ... Recall is defined as the ratio of the number of words correctly extracted to the total number of words that should have been extracted. Table7 , which shows the evaluation results in terms of precision and recall, indicates that SA is accurate in its extraction of keywords. (Although the evaluation cannot be viewed as completely objective, we can still observe certain tendencies in it.... ..."

### Table 2 Evaluation of the progressive algorithm

2003

"... In PAGE 8: ... Normally, the more accurately we could reconstruct the 3D object, the more accurate the sketch drawings became. Table2 shows the evaluation of the progressive algorithm for various projections and objects. It shows that the proposed algorithm recon- structs the most plausible objects from the sketch drawings in short time.... ..."

### Table 4: Rounding Modes for Accurate Expressions In practice, dot product expressions may contain a large number of terms making an explicit notation very cumbersome. To alleviate this di culty in mathematics, the symbol P is used. If for instance A and B are n-dimensional matrices, then the evaluation of

"... In PAGE 18: ... To obtain the unrounded or correctly rounded result of a dot product expression, the user needs to parenthesize the expression and precede it by the symbol # which may optionally be followed by a symbol for the rounding mode. Table4 shows the possible rounding modes with respect to the dot product expression form.... ..."