• Journal of the Korean Data and Information Science Society
• /
• v.20 no.6
• /
• pp.1191-1197
• /
• 2009

Most Bayesian approaches to model selection of wavelet analysis have drawbacks that computational cost is expensive to obtain accuracy for the fitted unknown function. To overcome the drawback, this article introduces loss functions which are criteria for level dependent threshold selection in wavelet based Bayesian methods with arbitrary size and regular design points. We demonstrate the utility of these criteria by four test functions and real data.

• Earthquakes and Structures
• /
• v.9 no.1
• /
• pp.257-279
• /
• 2015

The primary objective of this study is to summarize results from previous experimental tests on laboratory specimens of RC/steel frames with masonry infills, in order to develop fragility functions that permit the estimation of damage in typical non-structural components of RC frame buildings, as a function of attained peak interstory drift. The secondary objective is to derive loss functions for such non-structural components, which provide information on the probability of experiencing a certain level of monetary loss when a given damage state is attained. Fragility curves and loss function developed in this study can be directly used within the FEMA P-58 framework for the seismic performance assessment of RC frame buildings with masonry infills.

• The Journal of the Acoustical Society of Korea
• /
• v.40 no.2
• /
• pp.176-182
• /
• 2021

This paper evaluates and compares the performance of the Deep Nerual Network (DNN)-based speech enhancement models according to various loss functions. We used a complex network that can consider the phase information of speech as a baseline model. As the loss function, we consider two types of basic loss functions; the Mean Squared Error (MSE) and the Scale-Invariant Source-to-Noise Ratio (SI-SNR), and two types of perceptual-based loss functions, including the Perceptual Metric for Speech Quality Evaluation (PMSQE) and the Log Mel Spectra (LMS). The performance comparison was performed through objective evaluation and listening tests with outputs obtained using various combinations of the loss functions. Test results show that when a perceptual-based loss function was combined with MSE or SI-SNR, the overall performance is improved, and the perceptual-based loss functions, even exhibiting lower objective scores showed better performance in the listening test.

• Journal of the Korean earth science society
• /
• v.31 no.5
• /
• pp.437-447
• /
• 2010

Spatial estimation of environmental variables has been regarded as an important preliminary procedure for decision-making. A minimum variance criterion, which has often been adopted in traditional kriging algorithms, does not always guarantee the optimal estimates for subsequent decision-making processes. In this paper, a geostatistical framework is illustrated that consists of uncertainty modeling via stochastic simulation and risk modeling based on loss functions for the selection of optimal estimates. Loss functions that quantify the impact of choosing any estimate different from the unknown true value are linked to geostatistical simulation. A hybrid loss function is especially presented to account for the different impact of over- and underestimation of different land-use types. The loss function-specific estimates that minimize the expected loss are chosen as optimal estimates. The applicability of the geostatistical framework is demonstrated and discussed through a case study of copper mapping.

• Earthquakes and Structures
• /
• v.10 no.5
• /
• pp.1181-1212
• /
• 2016

Fragility and loss functions are developed to predict damage and economic losses due to earthquake loading in Reinforced Concrete (RC) structural components with smooth rebars. The attention is focused on external/internal beam-column joints and ductile/brittle weak columns, designed for gravity loads only, using low-strength concrete and plain steel reinforcing bars. First, a number of damage states are proposed and linked deterministically with commonly employed methods of repair and related activities. Results from previous experimental studies are used to develop empirical relationships between damage states and engineering demand parameters, such as interstory and column drift ratios. Probability distributions are fit to the empirical data and the associated statistical parameters are evaluated using statistical methods. Repair costs for damaged RC components are then estimated based on detailed quantity survey of a number of pre-70 RC buildings, using Italian costing manuals. Finally, loss functions are derived to predict the level of monetary losses to individual RC components as a function of the experienced response demand.

• Communications for Statistical Applications and Methods
• /
• v.14 no.2
• /
• pp.389-399
• /
• 2007

In this paper, Bayes estimates for the parameters k, c and reliability function of the Burr type XII model based on a type II censored samples under asymmetric loss functions viz., LINEX and SQUAREX loss functions are obtained. An approximation based on the Laplace approximation method (Tierney and Kadane, 1986) is used for obtaining the Bayes estimators of the parameters and reliability function. In order to compare the Bayes estimators under squared error loss, LINEX and SQUAREX loss functions respectively and the maximum likelihood estimator of the parameters and reliability function, Monte Carlo simulations are used.

• Journal of Korea Multimedia Society
• /
• v.17 no.9
• /
• pp.1098-1103
• /
• 2014

Smartphone is very much ingrained into our lives and it has propagated rapidly. With the growing expansion of smartphone, the number of lost smartphones has proliferated all the while. Unlike common personal belongings, smartphone is a high-priced goods and has personal information. In case it gets lost or stolen, it will cause monetary damage as well as secondary damage of personal information leaks. To lesson the damage, diverse applications that prevent the loss of a smartphone have been developed. But existing applications place emphasis on providing countermeasures that tracks the missing smartphones. Therefore, this paper describes the design and implementation of a smartphone application which provides precautionary functions to prevent the loss of a smartphone and countermeasures to track the missing smartphone. The implemented application prevents the loss of a smartphone using satefy zone and movement detection. In the event we lose a smartphone, it helps us recover a lost property using location tracking and remote control functions quickly and efficiently.

• The Journal of the Korea Contents Association
• /
• v.21 no.3
• /
• pp.616-625
• /
• 2021

Deep neural networks are an approximation method that approximates an arbitrary function to a linear model and then repeats additional approximation using a nonlinear active function. In this process, the method of evaluating the performance of approximation uses the loss function. Existing in-depth learning methods implement approximation that takes into account loss functions in the linear approximation process, but non-linear approximation phases that use active functions use non-linear transformation that is not related to reduction of loss functions of loss. This study proposes parametric activation functions that introduce scale parameters that can change the scale of activation functions and location parameters that can change the location of activation functions. By introducing parametric activation functions based on scale and location parameters, the performance of nonlinear approximation using activation functions can be improved. The scale and location parameters in each hidden layer can improve the performance of the deep neural network by determining parameters that minimize the loss function value through the learning process using the primary differential coefficient of the loss function for the parameters in the backpropagation. Through MNIST classification problems and XOR problems, parametric activation functions have been found to have superior performance over existing activation functions.

• Structural Engineering and Mechanics
• /
• v.88 no.1
• /
• pp.13-23
• /
• 2023

Seismic risk assessment studies are one of the most crucial instruments for mitigating casualties and economic losses. This work utilizes fragility curves to evaluate the seismic risk of single-story precast buildings, which are generally favored in Marmara's organized industrial zones. First, the precast building stock in the region has been categorized into nine sub-classes. Then, seven locations in the Marmara region with a high concentration of industrial activities are considered. Probabilistic seismic hazard assessments were conducted for both the soil-dependent and soil-independent scenarios. Subsequently, damage analysis was performed based on the structural capacity and mean fragility curves. Considering four different consequence models, 630 sub-class-specific loss curves for buildings were obtained. In the current study, it has been determined that the consequence model has a significant impact on the loss curves, hence, average loss curves were computed for each case investigated. In light of the acquired results, it was found that the loss ratio values obtained at different locations within the same region show significant variation. In addition, it was observed that the structural damage states change from serviceable to repairable or repairable to unrepairable. Within the scope of the study, 126 average loss functions were presented that could be easily used by non-experts in earthquake engineering, regardless of structural analysis. These functions, which offer loss ratios for varying hazard levels, are valuable outputs that allow preliminary risk assessment in the region and yield sensible outcomes for insurance activities.

• The Journal of Korea Robotics Society
• /
• v.18 no.3
• /
• pp.330-336
• /
• 2023

In this paper, a novel line-based simultaneous localization and mapping (SLAM) using a loss function of vanishing point measurements is proposed. In general, the Huber norm is used as a loss function for point and line features in feature-based SLAM. The proposed loss function of vanishing point measurements is based on the unit sphere model. Because the point and line feature measurements define the reprojection error in the image plane as a residual, linear loss functions such as the Huber norm is used. However, the typical loss functions are not suitable for vanishing point measurements with unbounded problems. To tackle this problem, we propose a loss function for vanishing point measurements. The proposed loss function is based on unit sphere model. Finally, we prove the validity of the loss function for vanishing point through experiments on a public dataset.