### Table 11: Analysis of residuals and the squared residuals plus their cleaned versions after smoothing for autoregression and ARCH-effects for the audit set.

2005

"... In PAGE 130: ... For these reasons, it is not a good idea to use the latter smooth for assessing the macro- scopic trends of the audit set. In Table11 , we summarized some numbers that assess the residuals of the smooths and their cleaned versions for the audit set. In this case the super smooth seems to have the best diagnostics to serve as the smooth to extract the trend from the audit set.... In PAGE 142: ... also showed the clear downward trend using smooth regression techniques. From the diagnostics summarized in Table11 , the super smooth is the one with the least structure in its residuals, so that one serves best as the trend indicator. In Figure 102, we depict this trend, with and without taking outliers into account.... ..."

Cited by 3

### Table 11: Analysis of residuals and the squared residuals plus their cleaned versions after smoothing for autoregression and ARCH-effects for the audit set.

"... In PAGE 128: ... For these reasons, it is not a good idea to use the latter smooth for assessing the macro- scopic trends of the audit set. In Table11 , we summarized some numbers that assess the residuals of the smooths and their cleaned versions for the audit set. In this case the super smooth seems to have the best diagnostics to serve as the smooth to extract the trend from the audit set.... In PAGE 139: ... The final analysis did not only show the rise that was predicted three times, but it also showed the clear downward trend using smooth regression techniques. From the diagnostics summarized in Table11 , the super smooth is the one with the least structure in its residuals, so that one serves best as the trend indicator. In Figure 102, we depict this trend, with and without taking outliers into account.... ..."

### Table 1: The Bayes factor for smoothed scale space model based on uniform priors for location parameters 01 and 02 and difierent priors for the scale parameter 0.

"... In PAGE 9: ... Next, a set of gamma priors with difierent super pa- rameters were considered for the scale parameter, 0. In order to study the possible efiect of these super-parameter values on the Bayes factor for a plausible range of their values, Table1 presents a selected set of these quantities. It is evident from the table that for all these choices of priors the null hypothesis of no signal is rejected, implying the existence of a signal in the image.... ..."

### Table 13.3: Running times for large models

2004

Cited by 2

### Table 3. Super capacitor parameters

"... In PAGE 5: ... Note that Vsuper(0) (the initial value of Vsuper) is equal to Videal(0) if Isuper(0) (the initial value of Isuper) is zero. Table3 lists the parameter ratings of single double-layer super capacitor [14]. Table 3.... ..."

### Table 1: Nine point summary statistics on a per counter basis and the same for all the function points together.

2005

"... In PAGE 13: ... Therefore, we further zoom in on the distribution of the func- tion point totals per counter. As a first quantitative indication, we summarize for each counter a nine-point summary statistic in Table1 . The abbreviations in Table 1 are self-explanatory, except sd which is short for standard deviation.... In PAGE 13: ... The in- and external counters are abbreviated with i and e with a subscript. From Table1 , we can see that function point counters i6, i8, i12, and i13, only rarely counted function points. We quantified now much more concrete how the data is skewed, for instance by the some- times large differences between the median and the mean.... In PAGE 13: ... So we calculate the empirical probability den- sity functions (PDFs) both on a per counter basis, and for the overall distribution. For the 4 counters spotted in Table1 , it is not a good idea to develop their probability... In PAGE 17: ... What could be the cause of this group difference, that was not visible at the individual level? To that end we carry out some more tests. Suppose we would throw out all the internal counters that counted only one or two projects (see Table1 ). In Figure 6 we depicted a comparison between the distributions of the rest of the internal counters versus the external ones.... In PAGE 100: ... The end values in both plots 1 and 2 are closer together, which is not a surprise, since further on in the time series there are less outliers. In Table1 0, we provide numerical details on the smoothing operations. The top half of this table contains six- point summaries of all individual smooth values.... In PAGE 103: ...1691 0.2059 Table1 0: Summaries of the values of the smoothing operations, all combinations of the smooths, and the first seven, excluding the wiggly robust median smooth. not to jump to conclusions.... In PAGE 130: ... For these reasons, it is not a good idea to use the latter smooth for assessing the macro- scopic trends of the audit set. In Table1 1, we summarized some numbers that assess the residuals of the smooths and their cleaned versions for the audit set. In this case the super smooth seems to have the best diagnostics to serve as the smooth to extract the trend from the audit set.... In PAGE 137: ...0003 0.1242 Table1 1: Analysis of residuals and the squared residuals plus their cleaned versions after smoothing for autoregression and ARCH-effects for the audit set. effect this implies that the maximal value at the beginning is taken, which is severely damped by the smoothing operation, and the minimal value implies the same but then at the end of the time series.... In PAGE 137: ...1691 0.2102 Table1 2: Summaries of the values of the smoothing operations, all combinations of the smooths, and the first seven, excluding the wiggly robust median for the audit set. To that end, we will provide in Table 12 some numbers.... In PAGE 142: ... also showed the clear downward trend using smooth regression techniques. From the diagnostics summarized in Table1 1, the super smooth is the one with the least structure in its residuals, so that one serves best as the trend indicator. In Figure 102, we depict this trend, with and without taking outliers into account.... ..."

Cited by 3