• Communications for Statistical Applications and Methods
• /
• v.6 no.2
• /
• pp.345-356
• /
• 1999

This paper deals with the problem of estimating regression coefficients in nonlinear regression model having random regressor. The sufficient conditions for consistency of the $L_1$-estimator with random regressor are given and discussed in this paper. An example is given to illustrate the application of the main results.

• Proceedings of the KIEE Conference
• /
• 1995.07b
• /
• pp.733-734
• /
• 1995

In this paper, a computational scheme for the model reduction problem using balanced realization is introduced, The scheme is illustrated by an example. The algorithm is based on the characterization of the solution to the model reduction problem.

• Communications for Statistical Applications and Methods
• /
• v.1 no.1
• /
• pp.10-19
• /
• 1994

We study the change-point problem in the context of system reliability models. The maximum likelihood estimators are obtained based on the Jelinski and Moranda model. To find the approximate distribution of the change-point estimator, we suggest of parametric bootstrap method in which the estimators are substituted in the assumed model. Through an example we illustrate the proposed method.

• Convergence Security Journal
• /
• v.21 no.2
• /
• pp.57-66
• /
• 2021

Deep neural networks (DNNs) provide excellent performance for image, speech, and pattern recognition. However, DNNs sometimes misrecognize certain adversarial examples. An adversarial example is a sample that adds optimized noise to the original data, which makes the DNN erroneously misclassified, although there is nothing wrong with the human eye. Therefore studies on defense against adversarial example attacks are required. In this paper, we have experimentally analyzed the success rate of detection for adversarial examples by adjusting various parameters. The performance of the ensemble defense method was analyzed using fast gradient sign method, DeepFool method, Carlini & Wanger method, which are adversarial example attack methods. Moreover, we used MNIST as experimental data and Tensorflow as a machine learning library. As an experimental method, we carried out performance analysis based on three adversarial example attack methods, threshold, number of models, and random noise. As a result, when there were 7 models and a threshold of 1, the detection rate for adversarial example is 98.3%, and the accuracy of 99.2% of the original sample is maintained.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.46 no.1
• /
• pp.55-67
• /
• 2023

Lightweight face recognition models, as one of the most popular and long-standing topics in the field of computer vision, has achieved vigorous development and has been widely used in many real-world applications due to fewer number of parameters, lower floating-point operations, and smaller model size. However, few surveys reviewed lightweight models and reimplemented these lightweight models by using the same calculating resource and training dataset. In this survey article, we present a comprehensive review about the recent research advances on the end-to-end efficient lightweight face recognition models and reimplement several of the most popular models. To start with, we introduce the overview of face recognition with lightweight models. Then, based on the construction of models, we categorize the lightweight models into: (1) artificially designing lightweight FR models, (2) pruned models to face recognition, (3) efficient automatic neural network architecture design based on neural architecture searching, (4) Knowledge distillation and (5) low-rank decomposition. As an example, we also introduce the SqueezeFaceNet and EfficientFaceNet by pruning SqueezeNet and EfficientNet. Additionally, we reimplement and present a detailed performance comparison of different lightweight models on the nine different test benchmarks. At last, the challenges and future works are provided. There are three main contributions in our survey: firstly, the categorized lightweight models can be conveniently identified so that we can explore new lightweight models for face recognition; secondly, the comprehensive performance comparisons are carried out so that ones can choose models when a state-of-the-art end-to-end face recognition system is deployed on mobile devices; thirdly, the challenges and future trends are stated to inspire our future works.

• ETRI Journal
• /
• v.34 no.6
• /
• pp.858-868
• /
• 2012

In this paper, stochastic models for the cross-correlation of multiple channels are established based on measurement data collected using a wideband multiple-input multiple-output relay Band Exploration and Channel Sounder system at 3.7 GHz. We propose models for the cross-correlation characteristics of large-scale parameters (LSPs) between two links, that is, the base station and mobile station (MS) link and the relay station and MS link. The LSPs include shadow fading, Rician K-factor, delay spread, angle spread of arrival, and angle spread of departure. Furthermore, models are established for the cross-correlation of the small-scale fading in the impulse responses of two links. The statistics of these model parameters are investigated as functions of geometrical features of the multilink. They are extracted from a large amount of cross-correlation observations, which are obtained in three measurement sites along more than one hundred measurement routes. These models can be used together with the standard single-link channel models for the generation of correlated components, for example, path clusters, in two separate channels.

• Transactions of the Korean Society of Pressure Vessels and Piping
• /
• v.12 no.2
• /
• pp.1-24
• /
• 2016

The original fracture criteria developed by Maxey/Kiefner for axial through-wall and surface-cracked pipes have worked well for many industries for a large variety of relatively low strength and toughness materials. However, newer line pipe steels have some unusual characteristics that differ from these older materials. One example is a test data that has demonstrated that X80 line-pipe with an axial through-wall-crack can fail at pressures about 30 percent lower than predicted with commonly used analysis methods for older steels. Thus, it is essential to review the currently available models and investigate the applicability of these models to newer high-strength line pipe materials. In this paper, the available models for predicting the failure behavior of axial-cracked pipes (through-wall-cracked and external surface-cracked pipes) were reviewed. Furthermore, the applicability of these models to high-strength steel pipes was investigated by analyzing limited full-scale pipe fracture initiation test results. Based on the analyzed results, the shortcomings of the available models were identified. For both through-wall and surface cracks, the major shortcomings were related to the characterization of the material toughness, which generally leads to non-conservative predictions in the J-T analyses. The findings in this paper may be limited to the test data that were consider for this study. The requisite characteristics of a potential model were also identified in the present paper.

• Journal of Korean Society for Quality Management
• /
• v.40 no.3
• /
• pp.347-358
• /
• 2012

Purpose: Given the mixture components - process variables experimental data, we propose the strategy to find the proper combined model. Methods: Process variables are factors in an experiment that are not mixture components but could affect the blending properties of the mixture ingredients. For example, the effectiveness of an etching solution which is measured as an etch rate is not only a function of the proportions of the three acids that are combined to form the mixture, but also depends on the temperature of the solution and the agitation rate. Efficient designs for the mixture components - process variables experiments depend on the mixture components - process variables model which is called a combined model. We often use the product model between the canonical polynomial model for the mixture and process variables model as a combined model. Results: First we choose the reasonable starting models among the class of admissible product models and practical combined models suggested by Lim(2011) based on the model selection criteria and then, search for candidate models which are subset models of the starting model by the sequential variables selection method or all possible regressions procedure. Conclusion: Good candidate models are screened by the evaluation of model selection criteria and checking the residual plots for the validity of the model assumption. The strategy to find the proper combined model is illustrated with examples in this paper.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.37 no.4
• /
• pp.231-238
• /
• 2014

Optimum lot size calculation for real world manufacturing environment has been focused since last few decades. Several extensions have been made to the basic economic order and production order quantity models to realize the possible practical situations in industry. However, focus on work-in-process inventory has been ignored relatively. This paper provides a comprehensive review of the models developed for group technology based manufacturing environment focusing on work-in-process inventory. Models have been extended from a perfect manufacturing conditions to an imperfect manufacturing situation considering rework, rejection and inspection. Optimum lot size has been evaluated using a simple algebraic optimization approach. Significant parameters are highlighted using sensitivity analysis for the developed models. Numerical example is used to illustrate the utilization of such models in day-to-day production setups and the impact of significant factors' variation on total cost and optimum lot size.

• International conference on construction engineering and project management
• /
• 2005.10a
• /
• pp.1181-1186
• /
• 2005

In many construction projects, progress and efficiency are hampered by poor communication of discipline-specific models. For example, architects use 2D or 3D CAD models and builders use CPM diagrams, Gantt charts, and spreadsheets to show their view of the project. Nowadays, advanced computer visualization tools, 4D CAD or VR, can show these disparate models to understand cross-disciplinary impacts of design and construction decisions. In Korea, several leading companies have tried to apply these tools to their pilot projects from the design phase to the maintenance phase. These companies have expected that more project stakeholders could understand a construction schedule more quickly and completely with 4D visualization than with the traditional construction management tools. However, modeling of the 4D CAD or VR can be quite time-consuming and expensive to generate manually and has therefore limited the spread and use of these models. In order to adopt widely those models in construction industry, the areas that those tools could support to take large benefits in diverse functional areas of construction management need to be analyzed. In this study, researchers analyze the usefulness and limitations of the 4D models and VR in the construction industry, develop the strategy of application priority, and improve the 4D modeling method.