### Table 3: Possible monitor points.

1994

"... In PAGE 9: ... Monitoring of specific events can either be activated globally, for each node in the network, or locally, for a specific node or range of nodes. Table3 summarizes the implemented monitor parameters. 7 Discussion To evaluate the current simulator implementation, work is in progress on a number of case studies of different par- allel applications.... ..."

Cited by 7

### Table 1 shows the results of the study. For each model, the optimal parameters and a measure of the model apos;s average error are presented. Average error provides a single measure of performance that can be used to compare models, it is defined below.

"... In PAGE 3: ... error: 32.62% Table1 : User model parameter estimation results DISCUSSION The average errors of models 2-4 is significantly less than that for model 1, with the most sophisticated (model 4) yielding an improvement of 9.85 and 15.... ..."

### Table 2 Optimal cost-sensitive monitoring policy based on actual solution quality Time-step

"... In PAGE 13: ...2. Assuming solution quality can be measured accurately by the run-time monitor (an unrealistic assumption in this case) and assuming a monitoring cost of 1, the dynamic programming algorithm described earlier computes the monitoring policy shown in Table2 . The number in each cell of Table 2 represents how much additional time to allocate to the algorithm based on the observed quality of the solution and the current time.... In PAGE 14: ...4.) The policy shown in Table2 was constructed by assuming the actual quality of an approximate solution could be determined by the run-time monitor. This is an unrealistic assumption because the quality of the current tour is defined with reference to the length of an optimal tour.... ..."

### Table 2 Optimal cost-sensitive monitoring policy based on actual solution quality Time-step

"... In PAGE 13: ...2. Assuming solution quality can be measured accurately by the run-time monitor (an unrealistic assumption in this case) and assuming a monitoring cost of 1, the dynamic programming algorithm described earlier computes the monitoring policy shown in Table2 . The number in each cell of Table 2 represents how much additional time to allocate to the algorithm based on the observed quality of the solution and the current time.... In PAGE 14: ...4.) The policy shown in Table2 was constructed by assuming the actual quality of an approximate solution could be determined by the run-time monitor. This is an unrealistic assumption because the quality of the current tour is defined with reference to the length of an optimal tour.... ..."

### Table 1: Advantages of aggressive dimensionality reduction Data Set Full Dimensional Optimal Quality Optimal Quality 1%-thresholding 1%-thresholding

2001

"... In PAGE 10: ... The rationale behind these methods is that any change in the nearest neighbor from the full dimensionality leads to loss of information; the rationale behind our approach is to be aggressive in removing the dimensions which have low co- herence as noise; thus, on an overall basis the aggressiveness of a dimensionality reduction process which uses the coher- ence probability of the dimensions may lead to very low precision with respect to the original data but much higher e ectiveness and coherence. In order to illustrate our point, we have indicated (in Table1 ) the prediction accuracy us- ing a 1%-thresholding technique in which only those eigen- values which are less than 1% of the largest eigenvalue are discarded. This prediction accuracy is typically very close to the full dimensional accuracy and is signi cantly lower than the optimal accuracy for all 3 data sets (as illustrated in the accuracy charts of Figures 5, 8, 11).... In PAGE 10: ... Thus, such a drastic reduction in dimensionality does not attempt to mirror the original nearest neighbors in the data; but rather improves their quality by removing the noise e ects in high dimensionality. It is also clear from Table1 that the opti- mal accuracy dimensionality is signi cantly lower than the 1%-thresholding method. In fact, the dimensionality for the 1%-thresholding method is quite close to the full dimension- ality.... ..."

Cited by 13

### Table 2: Optimal Points

"... In PAGE 7: ... After some experimention it was determined that im- proved performance could be obtained by adding elitism and linear scaling such that the probability of selecting the best design in each population was 4 times that of selecting an average design in the same population. For each method, the table lists the number of trials, out of 250, in which either the global #28#23G#29 or one of the local #28#23L#29 optima listed in Table2 was located. Obviously, all of those optima lie between points in the discretized space.... ..."

### Table 1 Development of particulate matter quality assessment in rivers in relation to increasing levels of monitoring sophistication12

2007

"... In PAGE 3: ... C15 Bottom deposits: present concentrations of pollutants (type i, iii) and past concentrations of pollutants in some cases (type ii, iii). Table1 shows a scheme for three levels of sophistication (levels A, B, C) of sediment analyses for bottom sediment and total suspended solids in rivers. With respect to basin-scale informa- tion, full cover of suspended particulate matter quality through- out flood stage and sediment cores at selected sites where continuous sediment may have occurred (both level C) is needed.... ..."

### Table 1 Defined monitoring states.

"... In PAGE 3: ... When MSR(PM) = 0, the associated process is unmarked. The state of the monitor with respect to MSR(PM) depends on the setting of MMCR(DMS) and MMCR(DMR) as defined by Figure 2 and Table1 . The setting of the MSR(PM) bit is not altered by the execution of a supervisor call (SVC) instruction, but is set to zero when any other interrupt is fielded, so the monitoring effect of the PM bit is preserved across calls to the kernel.... In PAGE 4: ... 22 counters monitor simultaneously: 20 counters for selectable events Two nonselectable events One cycle counter * One soft error counter Five events for each of FPU, FXU, ICU, SCU Five incrementors SCU monitors Organization of monitor hardware. Table1 shows the set of conditions that can be used to qualify the control of the performance monitor counters. Thus, with the MMCR the MSR can control the state of the monitor.... ..."