Results 1 - 10 of 90,660

Table 2. Classification error rates for simu- lated and real geon models

in Representing and Recognizing Complete Set of Geons Using Extended
by Superquadrics Lin Zhou, Lin Zhou, Ra Kambhamettu
"... In PAGE 4: ... Then, the selected 13 parameters are input to the classifiers to get their shape classes. Table2 shows the error rates for both simulated data and real data when using the nearest neighbor classi- fier and the BP neural network. As can be seen, the BP neu- ral network performs much better than the nearest neighbor classifier.... ..."
Cited by 1

Table 2: Statistical characteristics of real noise in a light Meyer trace and noise correspondingly simu- lated using the correlation distortion method.

in Improving wireless simulation through noise modeling
by Hyungjune Lee, Alberto Cerpa, Philip Levis 2007
"... In PAGE 5: ... With a small difference of the first-order PMF, this ap- proach achieves the sameness of auto-correlation between short-term noise data. Table2 shows the mean, standard deviation, skewness, and kurtosis. However, heavy-traffic 802.... ..."
Cited by 3

Table 2: Simulation intervals for SPEC95 programs (millions of instructions discarded / millions of committed instructions simu- lated) and miss rate in a direct-mapped 64K first-level data cache.

in Recovery Mechanism for Latency Misprediction
by Enric Morancho, Jose Maria Llaberia, Angel Olive 2001
"... In PAGE 9: ... To decide the amount of instructions to be discarded and to be simulated, we performed an analysis on temporal behaviour. Table2 shows the interval selected for each benchmark. 5.... ..."
Cited by 23

Table 2: Simulation intervals for SPEC95 programs (millions of instructions discarded / millions of committed instructions simu- lated) and miss rate in a direct-mapped 64K first-level data cache.

in Recovery Mechanism for Latency Misprediction
by Enric Morancho, Jose Mari'a Llaberi'a, A`ngel Olive'departament D'arquitectura De Computadors 2001
"... In PAGE 9: ... To decide the amount of instructions to be discarded and to be simulated, we performed an analysis on temporal behaviour. Table2 shows the interval selected for each benchmark. 5.... ..."
Cited by 23

Table 2: Simulation intervals for SPEC95 programs (millions of instructions discarded / millions of committed instructions simu- lated) and miss rate in a direct-mapped 64K first-level data cache.

in Recovery Mechanism for Latency Misprediction
by Enric Morancho, José María Llabería, Àngel Olivé 2001
"... In PAGE 9: ... To decide the amount of instructions to be discarded and to be simulated, we performed an analysis on temporal behaviour. Table2 shows the interval selected for each benchmark. 5.... ..."
Cited by 23

Table 5: Trace-Driven Simulation Parameters. This table shows the cluster sizes, time periods, and total reboots events we were able to simulate using a trace-driven methodology. We dropped several machines from each cluster because their logs were much shorter than the rest, and so would have shortened the simu- lated time too much.

in Improving Cluster Availability Using Workstation Validation
by Taliver Heath, Richard P. Martin, Thu D. Nguyen 2002
"... In PAGE 8: ... In choosing the time window we had to balance between maxi- mizing the cluster size and maximizing the time period (to observe steady state behaviors). Table5 gives the parameters for our trace- driven simulations. Figure 7 gives an close-up view of a portion (100 days) of the raw data that we selected for two clusters.... ..."
Cited by 27

Table 5: Trace-Driven Simulation Parameters. This table shows the cluster sizes, time periods, and total reboots events we were able to simulate using a trace-driven methodology. We dropped several machines from each cluster because their logs were much shorter than the rest, and so would have shortened the simu- lated time too much.

in Improving Cluster Availability Using Workstation Validation
by Taliver Heath, Richard P. Martin, Thu D. Nguyen
"... In PAGE 8: ... In choosing the time window we had to balance between maxi- mizing the cluster size and maximizing the time period (to observe steady state behaviors). Table5 gives the parameters for our trace- driven simulations. Figure 7 gives an close-up view of a portion (100 days) of the raw data that we selected for two clusters.... ..."

Table 1: Comparison of genetic algorithm and simu- lated annealing for three planes con ict

in Genetic Algorithms for solving Air Traffic Control conflicts
by Jean-marc Alliot, Hervé Gruber, Georges Joly, Marc Schoenauer

Table 2: Extreme values of estimation errors for the simu- lated line scanner image

in POSE ESTIMATION OF LINE CAMERAS USING LINEAR FEATURES. Young-ran Lee and Ayman Habib
by unknown authors
"... In PAGE 4: ... Those roads are used as control fea- tures, which are represented by natural/free-form curves rather than strict straight lines. Table2 summarizes esti- mation errors associated with the derived EOP by consid- ering the above mentioned linear feature. As one can see, we have correctly determined the EOP associated with that scene.... ..."

Table 1: Bandwidth estimates and IMSE values for Example 1. These are all means of 50 simu- lated samples each consisting of 100 observations.

in Bandwidth Selection for Kernel Conditional Density Estimation
by David M Bashtannyk, Rob J Hyndman 2001
"... In PAGE 14: ... These are all means of 50 simu- lated samples each consisting of 100 observations. Figure 4 about here Figure 5 about here Figure 6 about here The bandwidths and estimated IMSE obtained are given in Table1 . These are the means of 50 simulated samples each consisting of n=100 observations.... ..."
Cited by 4
Results 1 - 10 of 90,660