Erich L. Lehmann. Nonparametrics: Statistical Methods Based on Ranks. Holden-Day, Inc., San Francisco, CA, 1975.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....the Mann Whitney statistic, written as a function of Wilcoxon s rank sum statistic, with e k = nm 2nm ; i.e. the statistic MWW = i=1 R i ) i;j I X i Y j . Other important 6 statistics of this form are the Fisher Yates normal score test and the one in Van der Waerden test (see [7], p. 96) We shall show that l=1 e R l is a U statistic, if and only if e k = a( nm ) b for some constants a and b. Let R l : Then, changing the order of summation, one gets that n 1;m = l 1 R l 1 i=l 1 R l ] Note that if R l 1 = 0 then l 1 = 0. ....

E.L. Lehmann, Nonparametrics: Statistical Methods Based on Ranks, Holden-Day, Inc., San Francisco (1975). 10

....individual expert ranking on a set of five images. After collecting information from the experts, statistical methods are used to combine the results from a number of experts to compare consistency across the experts and to compare their results with the prioritizations produced by our algorithms [13]. This process, which is based on accepted statistical methods for combining and comparing rankings, provides a quantitative measure of the performance of our algorithms. V. Planning and Scheduling In addition to downlink data selection, the prioritized rock information can be used for ....

E. L. Lehmann, Nonparametrics: Statistical Methods Based on Ranks, San Francisco, California: Holden-Day, 1975.

....minimum and maximum values of non zero weighting coefficient was 5 and 42 respectively. The corresponding L 2 error based on 10,000 data points for 200 sample realisations was measured to be 0.0035 0.0030. Due to the highly asymmetric nature of the distribution of errors a Rank sum Wilcoxon test [16] is applied and shows that both error distributions for the full Parzen and RSDE estimators, at the 5 significance level, are identical. This is a somewhat satisfying result in that the accuracy of the RSDE is shown to be the same as the Parzen for this particular density function. The resulting ....

Lehmann, E.L. (1975) Nonparametric Statistical Methods Based on Ranks, New York: McGraw-Hill.

....from random variables with the same expected value. The mean di#erence test is used to compare the relative performance of the di#erent trading strategies. The tests were made with the Mathematica program [29] Other tests such as Wilcoxon s Signed Rank test [30] or the Kruskal Wallis Rank test [31] may also be used. Nonparametric tests such as these may be preferable due to their robustness, i.e. their insensitivity to small departures of the idealized assumptions, on e.g. underlying distributions, but are on the other hand less sensitive. 4.2 Data sets Four test sets with simulation ....

E. L. Lehmann. Nonparametric Statistical Methods Based on Ranks. McGraw-Hill, New York, 1975. 28

....independent samples for each system. A normal approximation is used aThis is generally less than the 30 document sets in the test data, since not all systems were evaluated on all document sets and summary sizes. when more than 25 paired samples are available or when ties in the ranks exist [Lehmann 1975]. Because of the correlation expected across different length summaries from the same document set, the analysis cannot be extended to the entire set of summaries produced by each peer. We note cases where the Wilcoxon signed rank statistic is significant at the 5 level by capitalizing the ....

E. L. Lehmann. Nonparamet- rics: Statistical Methods Based on Ranks. Holden and Day, San Francisco, 1975.

.... statistical tools such as Cramer s V or the contingency coe#cient C [5] that are both # based, the covariance and the correlation coe#cient statistics, the Q statistic [6] or also non parametric correlation coe#cients as the Spearman rank order correlation coe#cient or the Kendall s tau [7]. In this paper we use some mutual information based measures for the evaluation of dependence among outputs errors in a learning machine proposed in [10] The main idea behind the application of those measures of dependence consists in interpreting the dependence among the outputs as the common ....

E.L. Lehmann. Nonparametrics: Statistical Methods based on Ranks. Holden-Day, S.Francisco, 1975.

....independent samples for each system. A normal approximation is used 5 This is generally less than the 30 document sets in the test data, since not all systems were evaluated on all document sets and summary sizes. when more than 25 paired samples are available or when ties in the ranks exist [ Lehmann 1975 ] Because of the correlation expected across di erent length summaries from the same document set, the analysis cannot be extended to the entire set of summaries produced by each peer. We note cases where the Wilcoxon signed rank statistic is signi cant at the 5 level by capitalizing the ....

E. L. Lehmann. Nonparametrics: Statistical Methods Based on Ranks. Holden and Day, San Francisco, 1975.

....J. 1980) Practical Nonparametric Statistics, 2nd edition. John Wiley, New York. Fienberg, S. E. 1983) The Analysis of Cross Classified Categorical Data, 2nd edition. The MIT Press, Cambridge, MA. Fleiss, J. L. 1981) Statistical Methods for Rates and Proportions, 2nd edition. Wiley, New York. Lehmann, E.L. 1975) Nonparametrics: Statistical Methods Based on Ranks. Holden Day, San Francisco. References 107 Rosner, B. 1986) Fundamentals of Biostatistics. Duxbury Press, Boston. Snedecor, G. W. and Cochran, W. G. 1980) Statistical Methods, 7th edition. Iowa State University Press, Ames, Iowa. ....

....with regard to the presence of outliers in the data; that is, they are not affected very much by outliers. This is not the case for the classical F tests. You can find detailed discussions of the Kruskal Wallis and Friedman rankbased tests in a number of books on nonparametric tests; for example, Lehmann (1975) and Hettmansperger (1984) The KruskalWallis Rank Sum Test When you have a one way layout, as in the section Experiments With One Factor in the previous chapter, you can use the Kruskal Wallis rank sum test kruskal.testto test the null hypothesis that all group means are equal. We illustrate ....

Lehmann, E.L. (1975). Nonparametrics: Statistical Methods Based on Ranks.

.... statistical tools such as Cramer s V or the contingency coecient C [11] that are both 2 based, the covariance and the correlation coecient statistics, the Q statistic [20] or also non parametric correlation coecients as the Spearman rank order correlation coecient or the Kendall s tau [22]. In this paper we propose measures based on mutual information for evaluating the dependence among the outputs and the output errors of learning machines. Some of the main applications of mutual information to machine learning problems concern modeling of self organized systems and feature maps ....

E.L. Lehmann. Nonparametrics: Statistical Methods based on Ranks. Holden-Day, S.Francisco, 1975.

.... in the literature to accomplish ordinal MDS using a di erent objective function (however extensive comparisons were not performed) Extensions of the above objective function are possible, including a soft version (using sigmoidal functions) of Kendall s coecient of rank order correlation, # [Lehmann(1975)] Even though rank information (as opposed to quantitativevalues) can be used to de ne coordinates in shape space, experimental precision still remains an issue. If, for example, hemagglutination inhibition assayvalues which do not di er by a factor of four are essentially indistinguishable due ....

Lehmann E. \Nonparametrics: Statistical Methods Based on Ranks", Holden-Day (San Fransisco), 1975

....such pairs make up a nontrivial fraction of our data. One solution to this problem is to discard the tied pairs and to perform the sign test on the remaining data. A second solution is to count half the tied pairs as positive differences and half as negative differences. See [Hemelryk, 1952, Lehmann, 1975] for analyses of the two solutions. We choose the second solution here because it is more conservative. Our tied observations are almost certainly run times that are very close, and counting half of these pairs as positive and half as negative supports the null hypothesis, providing a ....

....the distribution of the pair differences is symmetric about zero. The 6 Ten of the remaining eleven problems are ties, and one problem is doubly censored. 10 alternate hypothesis is that the pair differences are slanted towards positive (or negative) values, in a sense made precise by Lehmann [Lehmann, 1975, page 157] Under the null hypothesis, we expect the sum of the ranks corresponding to the positive differences to be at least as large as the sum of the ranks corresponding to the negative differences. The p value is equal to the probability that sum of the positive ranks (denoted by T ) is ....

Lehmann, E. L. 1975. Nonparametrics: Statistical Methods Based on Ranks. Holden Day, San Fransisco. 15

No context found.

Erich L. Lehmann. Nonparametrics: Statistical Methods Based on Ranks. Holden-Day, Inc., San Francisco, CA, 1975.

No context found.

Lehmann, E. L. Nonparametrics: Statistical Methods Based on Ranks. Prentice Hall, rev. 1st edition, 1998.

No context found.

Lehmann, EL. (1975). " Nonparametrics: Statistical Methods Based on Ranks ". Holden-Day, San Francisco.

No context found.

E. L. Lehmann and H. j. M. D'Abrera. Nonparametrics: Statistical Methods Based on Ranks. Prentice Hall, 1998.

No context found.

E. L. Lehmann and H. J. M. D'Abrera, Nonparametrics: Statistical Methods Based on Ranks. Englewood Cliffs, NJ: Prentice-Hall, 1998.

No context found.

Lehmann, E.: Nonparametric Statistical Methods Based on Ranks. McGraw-Hill, New York NY (1975)

No context found.

E. L. Lehmann (1975), Nonparametrics: Statistical Methods based on Ranks. Hoden-Day, San Francisco.

No context found.

E. L. Lehmann (1975), Nonparametrics: Statistical Methods based on Ranks. Hoden-Day, San Francisco.

No context found.

E. L. Lehmann and H. J. M. D'Abrera, Nonparametrics: Statistical Methods Based on Ranks. Prentice Hall College Div, 1998.

No context found.

E. L. Lehmann and H. J. M. D'Abrera, Nonparametrics: Statistical Methods Based on Ranks. Englewood Cliffs, NJ: Prentice-Hall, 1998.

No context found.

E. L. Lehmann. Nonparametrics: Statistical methods based on ranks. Holden-Day, 1975.

No context found.

E. L. Lehmann. Nonparametrics: Statistical Methods Based on Ranks. McGraw-Hill, 1975.

No context found.

Lehmann, E.L. (1975) Nonparametrics: Statistical Methods Based on Ranks. Holden and Day, San Francisco, CA.

No context found.

LEHMANN, E.L., Nonparametrics: Statistical Methods Based on Ranks, Holden-Day Inc., San Francisco, California (1975).

First 50 documents

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC