Tools
Sorted by:
Try your query at:
 |
 |
 |
 |
 |
 |
Results 1 - 10
of
80,616
Table IV. Retrieval Performance of the Collection Promise Method as Implemented, Compared with the Estimated Best-Possible Performance of the Method
1999
Cited by 63
Table 3. Promising approaches
"... In PAGE 4: ...ng data skewness. If one step makes little difference (e.g., feature selection for DT), we just set No as default to save computation time. Table3 lists the 12 promising ap- proaches to tackle data skewness. The approaches in Table 3 are derived from bias analysis.... In PAGE 4: ... Table 3 lists the 12 promising ap- proaches to tackle data skewness. The approaches in Table3 are derived from bias analysis. We now further evaluate them through comparative experi- ments to investigate whether they can improve performance of classifiers for text classification, and which one is more appropriate for highly skewed data.... ..."
Table 2: This table depicts the most promising array de- signs in terms of performance-chip area tradeo for the IU Benchmark application. The data are shown graphically in Figure 13.
1997
"... In PAGE 8: ... Normalizing for array size, which designs give the best performance per chip area? Here the most promising designs in terms of cost-performance have been selected from Figure 12, normalized for array size, and plotted in Figure 13. The points are described further in Table2 . Since only the IU Benchmark has been plotted so far, none of the datapath selections contained oating point support.... ..."
Cited by 4
Table 4. Promise Heuristics: Search and Policy Adherence with MAC
2005
"... In PAGE 6: ... However, for problems tested in this work, much better performance was observed when this quantity was maximised. Results for both heuristics are shown in Table4 ; clearly min summed promisea128 is an anti-heuristic, with poor promise and fail-firstness. Another surprising re- sult is that the promise measure for max summed promisea128 is not especially high, although its fail-firstness is rather good (cf.... In PAGE 6: ... One explanation is that promisea128 in this case is confounded with forward degree. Because of this, maximum average promisea128 was also tested, but the outcome was essentially the same ( Table4 ). Thus, we see that strategies designed with a given policy in mind do not necessarily adhere well to that policy.... ..."
Cited by 3
Table 2 shows very promising results with our simple prototype.
2005
"... In PAGE 12: ... Another improvement in the construction of the And-Or graph consists in computing only the states that are reachable from the initial state. Table2 reports on the performance of the prototype when applied to various ex- amples of the literature: the Alternating Bit Protocol (ABP), and the Bounded Retransmission Protocol (BRP), on which we verify ve di erent properties [4]. Table 2 shows very promising results with our simple prototype.... In PAGE 12: ... Table2 . Results obtained on Intel Xeon 3Ghz with 4Gb of memory : S and E: number of states and edges of the graph ; C: number of channels; EEC: execution time (in second) of an implantation of EEC.... ..."
Cited by 2
Table 2: Performance measures (Asteroids indicate the factual or final assignment (i.e. the promised period and the final date of production); T* denominates the simulation run length) total under-
Table 2 : Music parameter settings for several emotions
2001
Cited by 6
Table 1: Initial PVM-SCI performance results While the above gures are promising, they do not represent true performance, because a fully op- erational PVM implementation would necessarily have to performing multiplexing, avoid blocking, and dynamically negotiate connections with multiple partners. Incorporating these facilities into the routines, however, signi cantly degrades performance, by a factor of 2 to 3, as shown in table 1 in the columns labelled Actual. While these gures are still a factor of about 8 better than the optimum of pvm psend() and pvm precv() over Ethernet, they do not yet re ect the true SCI potential. At the time of this 6
1995
Cited by 2
Table 3 Comparison of DSP/microcontrollers for some promising wireless sensing platforms
2007
"... In PAGE 7: ... The table does not incorporate some new platforms, such as the Philips Research CoolfluxDSP/HxH1200A1 based platform and the new Intel Imote. In Table3 some of the most promising platforms of Table 2 are compared with these novel platforms Table 2 Comparison of some miniature wireless sensor platforms.(see next two pages) ... In PAGE 10: ...From Table3 we can draw the following conclusions regarding the performance of the devices with respect to each other. Processing performance is highest with the Intel iMote, but we have to be aware of the fact that the device does not incorporate any sensor and it is already rather big and unpackaged.... ..."
Table 4: Throughput of both hand-coded and ESTEREL versions On the code size aspect, the Protocol Compiler produces a code which has quite the same size as the hand-code written in C language. The executable code of the receiver side is even smaller (4% on 200 Kbytes). Recent optimization work on the ESTEREL apos;s code generation phase show that a sensible reduction of the generated code size is possible without a loss of performance. This is very promising for our automated approach.
1995
Cited by 14
Results 1 - 10
of
80,616
