• Journal of the Korean Society of Systems Engineering
• /
• v.18 no.2
• /
• pp.75-93
• /
• 2022

Machine learning (ML) data-driven meta-model is proposed as a surrogate model to reduce the excessive computational cost of the physics-based model and facilitate the real-time prediction of a nuclear power plant's transient response. To forecast the transient response three machine learning (ML) meta-models based on recurrent neural networks (RNNs); specifically, Long Short Term Memory (LSTM), Gated Recurrent Unit (GRU), and a sequence combination of Convolutional Neural Network (CNN) and LSTM are developed. The chosen accident scenario is a control element assembly withdrawal at power concurrent with the Loss Of Offsite Power (LOOP). The transient response was obtained using the best estimate thermal hydraulics code, MARS-KS, and cross-validated against the Design and control document (DCD). DAKOTA software is loosely coupled with MARS-KS code via a python interface to perform the Best Estimate Plus Uncertainty Quantification (BEPU) analysis and generate a time series database of the system response to train, test and validate the ML meta-models. Key uncertain parameters identified as required by the CASU methodology were propagated using the non-parametric Monte-Carlo (MC) random propagation and Latin Hypercube Sampling technique until a statistically significant database (181 samples) as required by Wilk's fifth order is achieved with 95% probability and 95% confidence level. The three ML RNN models were built and optimized with the help of the Talos tool and demonstrated excellent performance in forecasting the most probable NPP transient response. This research was guided by the Systems Engineering (SE) approach for the systematic and efficient planning and execution of the research.

• Korean Journal of Computational Design and Engineering
• /
• v.14 no.6
• /
• pp.424-433
• /
• 2009

In order to improve productivity and efficiency of ship production process, production technology converged with Information Technology can be considered. Mid-term scheduling based on long-term schedule of ship building and execution planning based on short-term production schedule have an important role in ship production processes and techniques. However, data used in the scheduling are from the experiences of the past, cognitive, and often inaccurate, moreover the updates of the data by formatted documents are not being performed efficiently. This paper designs the tracking and monitoring system for the curved plates forming process with shop level. At first step to it, we redefine and analyze the curved plates forming process by using SysML. From the definition and analysis of the curved plates forming process, we design the system with respect to operational view considering operational environment and interactions between systems included and scenario about operation, and with respect to system view considering functionalities and interfaces of the system. In order to study the feasibility of the system designed, a prototype of the system has been implemented with 13.56 MHz RHD hardware and application software.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.3
• /
• pp.1100-1118
• /
• 2021

In a wide range of ML applications, the training data contains privacy-sensitive information that should be kept secure. Training the ML systems by privacy-sensitive data makes the ML model inherent to the data. As the structure of the model has been fine-tuned by training data, the model can be abused for accessing the data by the estimation in a reverse process called model inversion attack (MIA). Although, MIA has been applied to shallow neural network models of recognizers in literature and its threat in privacy violation has been approved, in the case of a deep learning (DL) model, its efficiency was under question. It was due to the complexity of a DL model structure, big number of DL model parameters, the huge size of training data, big number of registered users to a DL model and thereof big number of class labels. This research work first analyses the possibility of MIA on a deep learning model of a recognition system, namely a face recognizer. Second, despite the conventional MIA under the white box scenario of having partial access to the users' non-sensitive information in addition to the model structure, the MIA is implemented on a deep face recognition system by just having the model structure and parameters but not any user information. In this aspect, it is under a semi-white box scenario or in other words a gray-box scenario. The experimental results in targeting five registered users of a CNN-based face recognition system approve the possibility of regeneration of users' face images even for a deep model by MIA under a gray box scenario. Although, for some images the evaluation recognition score is low and the generated images are not easily recognizable, but for some other images the score is high and facial features of the targeted identities are observable. The objective and subjective evaluations demonstrate that privacy cyber-attack by MIA on a deep recognition system not only is feasible but also is a serious threat with increasing alert state in the future as there is considerable potential for integration more advanced ML techniques to MIA.

• Asian Pacific Journal of Cancer Prevention
• /
• v.16 no.12
• /
• pp.4881-4883
• /
• 2015

A population-based relationship between low vitamin D status and increased cancer risk is now generally accepted. However there were only few studies reported on prognostic impact. To determine the effect of low vitamin D on progression of breast cancer, we conducted a cross-sectional analysis of vitamin D levels and clinico-pathological characteristics in 200 cases of breast cancer diagnosed during 2011-2012 at the National Cancer Institute of Thailand. Vitamin D levels were measured by high-performance liquid chromatography (HPLC). Clinical and pathological data were accessed to examine prognostic effects of vitamin D. We found that the mean vitamin D level was $23.0{\pm}6.61ng/ml$. High vitamin D levels (${\geq}32ng/ml$) were detected in 7% of patients, low levels (<32 ng/ml) in 93% Mean vitamin D levels for stages 1-4 were $26.1{\pm}6.35$, $22.3{\pm}6.34$, $22.2{\pm}6.46$ and $21.3{\pm}5.42ng/ml$ respectively (P=0.016) and 24.1 and 21.3 ng/ml for lymph node negative and positive cases (P=0.006). Low vitamin D level (<32 ng/ml) was significantly found in majority of cases with advanced stage of the disease (P=0.036), positive node involvement (P=0.030) and large tumors (P=0.038). Our findings suggest that low and decreased level of vitamin D might correlate with progression and metastasis of breast cancer.

• Asian Pacific Journal of Cancer Prevention
• /
• v.15 no.2
• /
• pp.1031-1033
• /
• 2014

Aim: To investigate the effect of hyperglycemia and hyperinsulinemia on prostate cancer risk. Materials and Methods: This hospital based study was carried out using data retrieved from the register maintained in the Department of Biochemistry of a tertiary care hospital of Kathmandu, Nepal between $31^{st}$ December, 2011 and $31^{st}$ October, 2013. The variables collected were age, serum cholesterol, serum calcium, PSA, fasting blood glucose, serum insulin. Analysis was performed by descriptive statistics and testing of hypothesis using Excel 2003, R 2.8.0, Statistical Package for the Social Sciences (SPSS) for Windows Version 16.0 (SPSS Inc; Chicago, IL, USA) and the EPI Info 3.5.1 Windows Version. Results: Of the total 125 subjects enrolled in our present study, 25 cases were of PCa and 100 were healthy controls. The mean value of fasting plasma glucose was 95.5 mg/dl in cases of prostatic carcinoma and the mean value of fasting plasma insulin was $5.78{\mu}U/ml$ (p value: 0.0001*). The fasting insulin levels ${\mu}U/ml$ were categorized into the different ranges starting from ${\leq}2.75$, >2.75 to ${\leq}4.10$, >4.10 to ${\leq}6.10$, > $6.10{\mu}U/ml$. The maximum number of cases of prostatic carcinoma of fasting insulin levels falls in range of > $6.10{\mu}U/ml$. The highest insulin levels (> $6.10{\mu}U/ml$) were seen to be associated with an 2.55 fold risk of prostatic carcinoma when compared with fasting insulin levels of (< $2.75{\mu}U/ml$). Conclusions: Elevated fasting levels of serum insulin appear to be associated with a higher risk of prostate cancer.

• Journal of Convergence for Information Technology
• /
• v.9 no.8
• /
• pp.9-19
• /
• 2019

The effects of artificial intelligence technology is social science research as research on the impact on industry and changes in daily life, etc. This means that developing 'emotion AI' will prepare 'next-generation 3D-vector-sensitive AI'. This suggests the main keywords of the tertiary AI decision-making power. Particularly important results will be achieved because of the importance of current unethical learning and the implementation of decision-making systems that reflect ethical value judgments. This is a data based simulation, and required (1)Available data, (2)the technology for the goal of simulation. This takes into account the general content of the intended simulation based research. Currently, existing researches focus on meaningful research motivation, but this study presents the direction of technology. So, empirical analysis is consistent with the decision-making power of each country vs. new technology firms for AI on ehtic responsibility. As a result, there is a need for a concrete contribution and interpretation that can be achieved for the ethic Responsibility, on the technical side of AI / ML. In AI decision making, analytic power of human empathy should be included tech own trust.

• Communications for Statistical Applications and Methods
• /
• v.16 no.6
• /
• pp.971-978
• /
• 2009

Poisson generalized linear mixed models(GLMMs) have been widely used for the analysis of clustered or correlated count data. For the inference marginal likelihood, which is obtained by integrating out random effects is often used. It gives maximum likelihood(ML) estimator, but the integration is usually intractable. In this paper, we propose how to obtain the ML estimator via Laplace approximation based on hierarchical-likelihood (h-likelihood) approach under the Poisson GLMMs. In particular, the h-likelihood avoids the integration itself and gives a statistically efficient procedure for various random-effect models including GLMMs. The proposed method is illustrated using two practical examples and simulation studies.

• Journal of Korean Biological Nursing Science
• /
• v.21 no.2
• /
• pp.133-140
• /
• 2019

Purpose: The purpose of this study was to check the degree of residual microbial contamination after disinfection of reusable suction containers, used in an intensive care unit (ICU) and present basic data for efficient use through cost analysis in comparison to disposable suction containers. Methods: This study was conducted on 32 reusable suction containers used in an ICU on a selected specific day. After disinfection and washing, specimens were collected from the used containers and cultured to check for microbial contamination. Additionally, a comparative narrative study analyzes the cost of using reusable suction containers and disposable suction containers. Data were analyzed with the SPSS WIN 20.0 program using real numbers and percentage ${\chi}^2$-test. Results: As a result of the study, microorganisms were found in all samples where in 30 were gram-positive (62.5%) while 13 were gram-negative (27.1%). Based on level of contamination, microorganisms were less than 10CFU/ml in 18 samples (56.3%); 11-99CFU/ml in six samples (18.8%); and more than 100CFU/ml in eight samples (25%). Cost per day for a reusable suction container was $10,655+{\alpha}$ while cost per day for a disposable suction container was 10,666 won. Conclusion: This study found that reusable suction containers, even after disinfection, accounted for factors of potential infection as well as microbial contamination. So, disposable suction containers are superior in cost-effectiveness and highly efficient for use with infected patients.

• Geomechanics and Engineering
• /
• v.30 no.1
• /
• pp.11-26
• /
• 2022

One of the most important issues in tunneling, is the squeezing phenomenon. Squeezing can occur during excavation or after the construction of tunnels, which in both cases could lead to significant damages. Therefore, it is important to predict the squeezing and consider it in the early design stage of tunnel construction. Different empirical, semi-empirical and theoretical-analytical methods have been presented to determine the squeezing. Therefore, it is necessary to examine the ability of each of these methods and identify the best method among them. In this study, squeezing in a part of the Alborz service tunnel in Iran was estimated through a number of empirical, semi- empirical and theoretical-analytical methods. Among these methods, the most robust model was used to obtain a database including 300 data for training and 33 data for testing in order to develop a machine learning (ML) method. To this end, three ML models of Gaussian process regression (GPR), artificial neural network (ANN) and support vector regression (SVR) were trained and tested to propose a robust model to predict the squeezing phenomenon. A comparative analysis between the conventional and the ML methods utilized in this study showed that, the GPR model is the most robust model in the prediction of squeezing phenomenon. The sensitivity analysis of the input parameters using the mutual information test (MIT) method showed that, the most sensitive parameter on the squeezing phenomenon is the tangential strain (ε_θ^α) parameter with a sensitivity score of 2.18. Finally, the GPR model was recommended to predict the squeezing phenomenon in tunneling projects. This work's significance is that it can provide a good estimation of the squeezing phenomenon in tunneling projects, based on which geotechnical engineers can take the necessary actions to deal with it in the pre-construction designs.

• The Journal of the Korea Contents Association
• /
• v.9 no.8
• /
• pp.18-31
• /
• 2009

Improved reliability of microarray data and its reproducibility lead to recent increment in demand of data sharing and utilization among laboratories, but house-keeping and publicly opened microarray experimental data can hardly be accessed and utilized since they are in heterogeneous formats according to the various experimental methods and microarray platforms. In this paper, we propose a microarray sharing method which can easily retrieve and integrate microarray data from different experiment platforms, data formats, normalization methods, and analysis methods. Our system is based on web-service technology. The biologists of each site are able to search UDDI(Universal Description, Discovery, and Integration) registry, and download microarray data with common data structure of standard format recommended by MGED(Microarray Gene Expression Databases) society. The common data structure defined in this paper consists of IDF(Investigation Design Format), ADF(Array Design Format), SDRF(Sample and Relationship Format), and EDF(Expression Data Format). These components play role as templates to integrate microarray data with various structure and can be stored in standard formats such as MAGE-ML, MAGE-TAB, and XML Schema. In addition, our system provides advanced tools of automatic microarray data submitter and file manager to manipulate local microarray data efficiently.