### Table 2: Non-parametric approaches

1997

"... In PAGE 2: ...Table 2: Non-parametric approaches The object coordinates of the control points have to be reduced to 2D-coordinates on the projection surface first. Table2 shows the necessary control points for non- parametric approaches, depending on the number of needed coefficients. 3.... ..."

Cited by 5

### Table 1: Difference between parametric and non-parametric approaches

2004

"... In PAGE 6: ... vi LIST OF TABLES Table Page Table1 : Difference between parametric and non-parametric approaches .... In PAGE 17: ... In statistical decision theoretical approach the decision boundaries are determined by probability distributions of patterns belonging to different classes. Table1 summarizes the differences in between parametric and non-parametric approaches: ... ..."

### Table 1: A comparison between parametric and non-parametric approaches.

### Table 2.14: Characteristics of Density Estimation Approach Non-Parametric Approach Gaussian GMM

2004

### TABLE I. COMPARISON OF EXISTING PARAMETRIC APPROACHES FOR MODELING FATIGUE AND PERFORMANCE WITH NONPARAMETRIC AND SEMIPARAMETRIC APPROACHES.

### Table 2.12: The mean squared error as deviation measure from the real data for the different non-parametric approaches.

in EURANDOM

2006

### Table 4 Upper Prediction Bounds for the Probability of a Space Shuttle Software Critical Failure within a Future Time Interval of 200 Test Hours Nonparametric Approach

"... In PAGE 7: ... However, the residuals produced by these models are significantly nonnormal (using the Shapiro-Wilk test of normality), and so the use of the parametric upper prediction bounds of Table 3 is inappropriate. Table4 presents results for the alternative nonparametric prediction bounds discussed in Section 4 that avoid this problem. Table 4 contains predictions of I+1,c p for the same 4 standard models as in Table 3 as well as for the best 1 variable model, in this case, the one with x=n-1 and an intercept.... In PAGE 7: ... Table 4 presents results for the alternative nonparametric prediction bounds discussed in Section 4 that avoid this problem. Table4 contains predictions of I+1,c p for the same 4 standard models as in Table 3 as well as for the best 1 variable model, in this case, the one with x=n-1 and an intercept. This model is best in the sense that it has minimal associated asymmetric half-width AHWc(M), as defined in Section 4, within the family of time and failure models.... In PAGE 7: ...These results indicate there is a noticeable benefit to consideration of the full family of time and failure models. Moreover, the nonparametric approach of Table4 does not require a normality assumption to justify its use while generating a 90% prediction bound that is over 14% smaller than the one generated by the parametric approach. While the parametric approach generates a conservative upper bound when applied to the space shuttle critical failure process, it is not clear that this will hold in general.... ..."

### Table 1. Significant parameters in the RQA approach. The statistical analysis was based on the two-sided t-test as well the non-parametric Mann-Whitney-U-test.

### Table 2: Comparison of raw algorithm to enhanced algorithm

2004

"... In PAGE 6: ...able 1: Difference between parametric and non-parametric approaches .....................7 Table2 : Comparison of raw algorithm to enhanced algorithm.... ..."

### lable to the N gt; 1 case when all the neighbors have the same marginal statistics, which, in practice requires they all belong to the same subband. Second, it is estimated from the noise-free coefficients, and it is difficult to extend it for use in the noisy case. We have also investigated a more direct maximum likeli- hood approach for estimating a nonparametric pz(z) from an observed set of neighborhood vectors:

2003

Cited by 106