Results 1 - 10 of 24,020

Table 14. Performance of various feature combinations on the task of argument identification.

in Support Vector Learning for Semantic Argument Classification
by Sameer Pradhan, Kadri Hacioglu, Valerie Krugler, Wayne Ward, James H. Martin, Daniel Jurafsky
"... In PAGE 17: ... Table14 shows the feature salience on the task of argument identification. As opposed to the argument classification task, where removing the path has the least effect on perfor- mance, on the task of argument identification, removing the path causes the convergence in SVM training to be very slow and has the most detrimental effect on the performance.... ..."

Table 2: Effect of SQR method on TREC-4 ad hoc queries truncated to various numbers of terms.

in V-Twin: A Lightweight Engine for Interactive Use
by Daniel E. Rose, Curt Stevens 1996
"... In PAGE 4: ... Then we ran the system with and without SQR. The results for the TREC-4 collection are shown in Table2 . We expected that SQR would improve precision for the shortened queries, and it did: 21.... In PAGE 10: ... These results are shown in Tables 8 and 9. As with our previous SQR experiment summarized in Table2 , we found again that SQR did not have a detrimental effect. In fact, it again improved performance, with the largest gains generally coming, as expected, on the most truncated (i.... ..."
Cited by 19

Table 1: Benchmark Equilibrium Outcomes and Sensitivity Analysis without Peer-Group Effects

in unknown title
by unknown authors 2002
"... In PAGE 21: ... In this sense, we may view 0 as a parameter that emphasizes the detrimental health effect while highlights the (conventional) addiction effect associated with past consumption of the addictive good. (Insert Table1 about here) The distinct effects are clearly present in our numerical results. We begin by describing the general equilibrium implications of the addictive effect.... ..."

Table 5 Mean SNR for all test matches

in Automatic Transcription of Football Commentaries in the MUMIS Project
by Janienke Sturm, Judith M. Kessens, Mirjam Wester, Febe de Wet, Eric Sanders, Eric S, Helmer Strik
"... In PAGE 3: ...Table5 shows that the SNRs are rather low (except for the German YugNet match). The SNR is lowest for the German EngGer match, which may explain the high baseline WER that was reported for this match (Table 3).... In PAGE 4: ...4. EXP4: Noise robust techniques It is clear that the presence of stadium noise in the speech signals (see Table5 ) has a detrimental effect on the performance; therefore we applied a number of noise robust techniques in an effort to alleviate the detrimental effect of the noise. Various noise robust techniques were studied: Cepstral Mean Normalisation (CMN) [7], Mean Variance Normalisation (MVN) [8], Histogram Normalisation (HN) [9][10], and a time-domain noise reduction (TDNR) technique [11].... ..."

Table 2: Relationship between Local Minima and Detrimental Crossover

in Statistical exploratory analysis of genetic algorithms
by Andrew Czarn, Cara Macnish, Kaipillil Vijayan, Berwin Turlach 2004
"... In PAGE 14: ... Chromosomes in these local minima were more likely to be shifted towards the global minimum due to the bias of the mutation operator. Overviewing the results for all the rotated functions, we observed that if roughly 20% or more of the local minima lay along the x or y axes, as shown in Table2 , the crossover operator proved to be beneflcial for the function, otherwise it was detrimental. More generally speaking, this axial bias is a special case of the more general relationship between the problem encoding and the solution space, discussed below.... ..."
Cited by 9

Table 1. Summary of Leakage Reduction Techniques Technique Benefit Detriment

in Static Energy Reduction Techniques for Microprocessor Caches
by Heather Hanson, M.S. Hrishikesh, Vikas Agarwal, Stephen W. Keckler, Doug Burger 2001
Cited by 25

Table 3 Speedup and E ciency of detrimental anomaly. P T

in Distributed Branch And Bound Algorithms For Global Optimization
by Ioannis P. Androulakis, Christodoulos A. Floudas 1998
Cited by 3

Table 4: Continuous multi-dimensional experiments: Scal- ing in dimension D2 as proposal distribution C9 is widened. Distributions are multivariate Gaussians. 1000 repetition- s with estimation samples of size 200. AM C8

in
by unknown authors
"... In PAGE 8: ... We also ran a series of experiments that varied the width of the proposal distribution C9. Table4 shows that weakening C9 damaged the performance of importance sampling (pre- dictably), while the greedy search seemed to mitigate the effects of a poor C9 to the extent that its detrimental effects are significantly diminished. Next,toscaleuptoaslightlymorecomplextaskwecon- sidered a distribution C8 that was a mixture of Gaussians while keeping C9 unimodal.... ..."

Table 2: The payoff matrix for Player One in Rock-Paper- Scissors.

in Safe Strategies for Agent Modelling in Games
by unknown authors
"... In PAGE 4: ... These experiments show that epsilon1-safe strategies are able to improve on all but the best opponent modellers, without a detrimental effect on the modellers that could not be improved. Rock-Paper-Scissors Rock-Paper-Scissors (RPS) is a simple two-player, zero-sum matrix game, with the payoff matrix defined in Table2 . The optimal game theoretic strategy in RPS is to play all actions with equal probability.... ..."

Table 5: A comparison of safe resynthesis, unsafe resynthesis, and unsafe followed by safe resynthesis. Relative increases in route length and via count are shown.

in SafeResynth: A New Technique for Physical Synthesis
by Kai-hui Chang, Igor L. Markov, Valeria Bertacco 2008
"... In PAGE 19: ... In addition, performing safe optimiza- tions avoids the detrimental effects that worsen other physical parameters. As can be observed from Table5 , performing safe instead of unsafe resynthesis avoids the signi cant increase in via count. The results also suggest that to obtain the greatest improvement, the advantages of both safe and unsafe techniques should be leveraged.... ..."
Results 1 - 10 of 24,020