• Communications for Statistical Applications and Methods
• /
• v.2 no.2
• /
• pp.52-62
• /
• 1995

Consider an estimation problem under squared error loss in a family of non-regular densities with both terminals of the support being decreasing functions of an unknown parameter. Using Karlin's(1958) technique, sufficient conditions are given for generalized Bayes estimators to be admissible for estimating an arbitrarily positive, monotone parametric function and then treat some examples which illustrate our results.

• IE interfaces
• /
• v.7 no.1
• /
• pp.75-80
• /
• 1994

In a recent article, Leon et al. lucidly explained the ideas of the Taguchi two-stage procedure for parameter design optimization, and proposed alternative performance measures called PerMIA to the signal-to-noise ratios. On the other hand, Box proposed an empirical approach to the problem based upon monotone transformations of the performance characteristic(y). This paper develops procedures for parameter design optimization under the assumptions that the expected loss(not necessarily a mean squared error loss) is increasing with respect to the variance of the error in y, and that the mean of y satisfies certain conditions of adjustability. It turns out that the variance of the error in y can play the role of PerMIA, and it is further shown that the derived PerMIA can be adapted to the Box empirical procedure for the minimization of the expected loss in the original metric.

• Journal of the Korean Data and Information Science Society
• /
• v.28 no.5
• /
• pp.1205-1215
• /
• 2017

In this paper, we aim to estimate two scale-parameters of exponentiated Pareto distribution (EPD) based on lower record values. Record values arise naturally in many real life applications involving data relating to weather, sport, economics and life testing studies. We calculate the Bayesian estimators for the two parameters of EPD based on lower record values. The Bayes estimators of two parameters for the EPD with lower record values under the squared error loss (SEL), linex loss (LL) and entropy loss (EL) functions are provided. Lindley's approximate method is used to compute these estimators. We compare the Bayesian estimators in the sense of the bias and root mean squared estimates (RMSE).

• International Journal of Quality Innovation
• /
• v.4 no.1
• /
• pp.191-204
• /
• 2003

The target costing technique, mathematically discussed by Sauers, only uses the $C_p index along with Taguchi loss function and $\bar{X}$-P control charts to setup goal control limits. The new specification limits derived from Taguchi loss function is linked through the $C_p value to $\bar{X}$-P control charts to obtain goal control limits. Studies have shown that the point estimator of the $C_p index, $C_p, could vary from time to time due to the sampling error. The suggested approach is to use confidence intervals, especially the lower confidence intervals, to replace the point estimator. Therefore, an improvement on target costing technique is presented by applying the lower confidence interval of the $C_p index and using both Taguchi and Spiring's loss functions together with $\bar{X}$-P charts to make this technique more robust in practice. An example is also provided to illustrate how the improved target costing technique works.

• The Korean Journal of Applied Statistics
• /
• v.7 no.2
• /
• pp.131-143
• /
• 1994

Under the asymmetric losses (entropy loss and Stein loss), we find the classes of Bayes and empiricla Bayes estimates for estimating the Poisson means when the distributin of means are believed a priori. Following the idea of Efron and Morris (1973), we have a computer simulation to compute a relative savings loss of proposed estimates as compared to the classical estimates.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.16 no.5
• /
• pp.179-186
• /
• 2021

For stereo matching based on deep learning, the design of network structure is crucial to the calculation of matching cost, and the time-consuming problem of convolutional neural network in image processing also needs to be solved urgently. In this paper, a method of stereo matching using sparse loss volume in parallax dimension is proposed. A sparse 3D loss volume is constructed by using a wide step length translation of the right view feature map, which reduces the video memory and computing resources required by the 3D convolution module by several times. In order to improve the accuracy of the algorithm, the nonlinear up-sampling of the matching loss in the parallax dimension is carried out by using the method of multi-category output, and the training model is combined with two kinds of loss functions. Compared with the benchmark algorithm, the proposed algorithm not only improves the accuracy but also shortens the running time by about 30%.

• KIPS Transactions on Software and Data Engineering
• /
• v.11 no.9
• /
• pp.371-380
• /
• 2022

Convolutional neural networks are widely used to manipulate data arranged in a grid, such as images. A general convolutional neural network consists of a convolutional layers and a fully connected layers, and each layer contains a nonlinear activation functions. This paper proposes a combined parametric activation function to improve the performance of convolutional neural networks. The combined parametric activation function is created by adding the parametric activation functions to which parameters that convert the scale and location of the activation function are applied. Various nonlinear intervals can be created according to parameters that convert multiple scales and locations, and parameters can be learned in the direction of minimizing the loss function calculated by the given input data. As a result of testing the performance of the convolutional neural network using the combined parametric activation function on the MNIST, Fashion MNIST, CIFAR10 and CIFAR100 classification problems, it was confirmed that it had better performance than other activation functions.

• Geomechanics and Engineering
• /
• v.37 no.6
• /
• pp.565-576
• /
• 2024

In this study, the earthquake risk assessment of single-story RC precast buildings in Turkey was carried out using loss curves. In this regard, Kocaeli, a seismically active city in the Marmara region, and this building class, which is preferred intensively, were considered. Quality and period parameters were defined based on structural and geometric properties. Depending on these parameters, nine main sub-classes were defined to represent the building stock in the region. First, considering the mean fragility curves and four different central damage ratio models, vulnerability curves for each sub-class were computed as a function of spectral acceleration. Then, probabilistic seismic hazard analyses were performed for stiff and soft soil conditions for different earthquake probabilities of exceedance in 50 years. In the last step, 90 loss curves were derived based on vulnerability and hazard results. Within the scope of the study, the comparative parametric evaluations for three different earthquake intensity levels showed that the structural damage ratio values for nine sub-classes changed significantly. In addition, the quality parameter was found to be more effective on a structure's damage state than the period parameter. It is evident that since loss curves allow direct loss ratio calculation for any hazard level without needing seismic hazard and damage analysis, they are considered essential tools in rapid earthquake risk estimation and mitigation initiatives.

• Journal of Periodontal and Implant Science
• /
• v.45 no.6
• /
• pp.216-222
• /
• 2015

Purpose: After extraction, the alveolar bone tends to undergo atrophy in three-dimensions. The amount of alveolar bone loss in the horizontal dimension has been reported to be greater than the amount of bone loss in the vertical dimension, and is most pronounced in the buccal aspect. The aim of this study was to monitor the predictive alveolar bone level following the extraction of anterior teeth seriously involved with advanced chronic periodontitis. Methods: This study included 25 patients with advanced chronic periodontitis, whose maxillary anterior teeth had been extracted due to extensive attachment loss more than one year before the study. Periapical radiographs were analyzed to assess the vertical level of alveolar bone surrounding the edentulous area. An imaginary line connecting the mesial and the distal ends of the alveolar crest facing the adjacent tooth was arbitrarily created. Several representative coordinates were established in the horizontal direction, and the vertical distance from the imaginary line to the alveolar crest was measured at each coordinate for each patient using image analysis software. Regression functions predicting the vertical level of the alveolar bone in the maxillary anterior edentulous area were identified for each patient. Results: The regression functions demonstrated a tendency to converge to parabolic shapes. The predicted maximum distance between the imaginary line and the alveolar bone calculated using the regression function was $1.43{\pm}0.65mm$. No significant differences were found between the expected and actual maximum distances. Likewise, the predicted and actual maximum horizontal distances did not show any significant differences. The distance from the alveolar bone crest to the imaginary lines was not influenced by the mesio-distal spans of the edentulous area. Conclusions: After extraction, the vertical level of the alveolar ridge increased to become closer to the reference line connecting the mesial and distal alveolar crests.

• Journal of the Korean GEO-environmental Society
• /
• v.16 no.10
• /
• pp.15-22
• /
• 2015

In this study, to evaluate the long-term behavior characteristics and the loss of prestress force, the long-term measurement of the tensile anchors in the actual construction was performed and the results were analyzed comparing with the existing estimation. As the reinforcement member used for the purpose of slope stability or uplift-resisting of the permanent structure, etc, the permanent anchor should maintain the functions during the performance period of the structure differently from the temporary anchor. However, as the time passes by, since the relaxation and the creep of the anchor occur constantly, the management for the loss of tensile force is essential to perform the functions stably. So far, the loss of the tensile force has been estimated according to the reduction of the prestress using elasticity theory and using the relaxation value according to the type of tension member and the test using the long-term measurement is limited. Therefore, in this study, the site condition and the ground were investigated for the tensile anchor in the actual construction and the long-term measurement results more than 500 days was analyzed by installing the loadcell, inclinometer and the groundwater level gauge. In addition, the long-term behavior characteristics were evaluated by comparing the disposition of the measured earth retaining wall and the tension force loss of the anchor with the existing interpretation results. In the evaluation results, the most of the tension force loss occurs within 90 days and the loss was measured less than the estimated values.