R.V. Hogg, "Adaptive robust procedure: A partial review and some suggestions for the future," Journal of American Statistical Association, vol.69, pp.909-923, 1974. Home/Search Document Not in Database Summary Related Articles Check |
This paper is cited in the following contexts:
Object Classification in 3-D Images Using Alpha-Trimmed Radial.. - Bors, Pitas (1999) (Correct)
.... a good location estimate in the case when data is contaminated with medium and long tailed distributions [31] and has been employed for image filtering [11, 32] In the approach considered in this study, the number of samples trimmed away from the given distribution depends on the data statistics [33]. The classical RBF and MRBF training algorithms are particular cases of the proposed algorithm. We prove that Alpha Trimmed Mean RBF algorithm provides unbiased estimates for the ellipses center and this result can be extended for ellipsoids. A trimming algorithm which discards data samples with ....
....assigned to the k th basis function at the moment t. We can observe that for a=0 we obtain the LVQ algorithm for 7 minimum output variance [39] The parameter a is chosen according to the data distribution. The following measure is used for estimating the tail length of the data distribution [33]: Q = 13) where U[ L[ represent the average of the upper and respectively the lower percent of data samples assigned to a specific basis function. The number of data samples to be trimmed away relies directly on the value of Q: i Q (14) 2 For long tailed distributions, the amount ....
[Article contains additional citation context not shown here]
R. V. Hogg, "Adaptive robust procedures: a partial review and some suggestions for future applications and theory," J. Am. Stat. Assoc., no. 348, vol. 69, pp. 909-923, 1974.
Robust RBF Networks - Bors, Pitas (Correct)
....samples assigned to the # th basis function at the moment #. We can observe that for # # =0 we obtain the LVQ algorithm for minimum output variance [29] The parameter # # is chosen according to the data distribution. The following measure is used for estimating the tail of the data distribution [35], 36] 19) where # ###, #### represent the average of the upper and respectively the lower # percent of data samples assigned to a specific basis function. The number of data samples to be trimmed away relies directly on the value of Q: # (20) For long tailed distributions, the amount ....
Hogg, R.V. (1974), "Adaptive robust procedures: a partial review and some suggestions for future applications and theory," J. Am. Stat. Assoc., vol. 69, no. 348, pp. 909-923.
Biomedical Image Processing with Morphology-Based Nonlinear Filters - Schulze (1994) (Correct)
....[53] The resulting overall noise process is often described as heavy tailed Gaussian, meaning that the noise is roughly additive Gaussian noise with some additional impulses present. However, the noise present in thermal images is not overwhelmingly impulsive, as will be shown below. Hogg [64] and David [25] defined a statistic q that is a good indicator of the tail length of a distribution. This statistic is given in equation (5.1) below [25] q = u 0.05 ( l 0.05 ( u 0.5 ( l 0.5 ( where u b ( mean of the largest nb order statistics of the distribution l b ( ....
....a high value of q indicates that the noise process is impulsive. Ryu estimated the value of q for the noise process in thermal images to be q 2.32, compared to a value for additive Gaussian noise ofq 2.19[53] However, Hogg quotes a value of q = 2. 58 for the normal (Gaussian) distribution [64]. Ryu indicated that the thermal image noise process is slightly more impulsive than Gaussian noise, but the results of Hogg indicate that it is slightly less impulsive. Regardless, the thermal image noise process is not very impulsive, but is actually very similar to Gaussian noise. The ....
[Article contains additional citation context not shown here]
R. V. Hogg, "Adaptive robust procedures: A partial review and some suggestions for future applications and theory," J. Amer. Statist..Assoc., vol. 69, no. 348, pp. 909-923, 1974.
Object Classification in 3-D Images Using Alpha-Trimmed Mean.. - Bors, Pitas (Correct)
.... a good location estimate in the case when data is contaminated with medium and long tailed distributions [31] and has been employed for image ltering [11, 32] In the approach considered in this study, the number of samples trimmed away from the given distribution depends on the data statistics [33]. The classical RBF and MRBF training algorithms are particular cases of the proposed algorithm. We prove that Alpha Trimmed Mean RBF algorithm provides unbiased estimates for the ellipses center and this result can be extended for ellipsoids. A trimming algorithm which discards data samples with ....
....to the k th basis function at the moment t. We can observe that for k =0 we obtain the LVQ algorithm for 7 minimum output variance [39] The parameter k is chosen according to the data distribution. The following measure is used for estimating the tail length of the data distribution [33] : Q = U[0:5] L[0:5] U[0:05] L[0:05] 13) where U [ L[ represent the average of the upper and respectively the lower percent of data samples assigned to a speci c basis function. The number of data samples to be trimmed away relies directly on the value of Q : k = 1 Q 2 : 14) ....
[Article contains additional citation context not shown here]
R. V. Hogg, \Adaptive robust procedures: a partial review and some suggestions for future applications and theory," J. Am. Stat. Assoc., no. 348, vol. 69, pp. 909-923, 1974.
Comparing The Sample Mean And The Sample Median: An Exploration.. - Sherman (Correct)
....statistics. We note that the exponential power family has been discussed by, e.g. Box and Tiao (1973) although their formulation only allows for p 2(1,1 ) The properties of various robust and adaptive estimators for data from this family have been studied by, e.g. D Agostino and Lee (1974) and Hogg (1974). Also we note that the estimators, x and x, are considered here for ease of exposition. These are the maximum likelihood estimators only when p=2 and p=1, respectively. For other known values of p, the asymptotically best linear estimator can be obtained using the results of Chernoff, ....
Hogg, R.V. (1974). Adaptive Robust Procedures: A Partial Review and Some Suggestions for the Future, Journal of the American Statistical Association, 69, 909-923.
Fixed-Point ICA Based Speech Signal Separation and Enhancement.. - Prasad (2005) (Correct)
No context found.
R.V. Hogg, "Adaptive robust procedure: A partial review and some suggestions for the future," Journal of American Statistical Association, vol.69, pp.909-923, 1974.
Robustness Properties of Johns' Estimators - Peter Mcburney Mathematics (Correct)
No context found.
R. V. Hogg. Adaptive robust procedures: a partial review and suggestions for future applications and theory. Journal of the American Statistical Association, 69:909--927, 1974.
LS-SVM Regression Modelling and its Applications - De Brabanter (2004) (Correct)
No context found.
Hogg, R.V. (1974). Adaptive robust procedures: A partial review and some suggestions for future applications and theory. J. Amer. Statist. Assoc. , 348, 909-927.
Nonparametric Estimation of Weighted Proportions With.. - Goldenshluger, Faraggi (Correct)
No context found.
Technometrics 12, 55--67. Hogg, R. (1974). Adaptive robust procedures: a partial review and some suggestions for future applications and theory. J. Am. Statist. Assoc. 69, 909--27.
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