• ETRI Journal
• /
• v.41 no.6
• /
• pp.771-781
• /
• 2019

We propose the VS³-NET model to solve the task of question answering questions with machine-reading comprehension that searches for an appropriate answer in a given context. VS³-NET is a model that trains latent variables for each question using variational inferences based on a model of a simple recurrent unit-based sentences and self-matching networks. The types of questions vary, and the answers depend on the type of question. To perform efficient inference and learning, we introduce neural question-type models to approximate the prior and posterior distributions of the latent variables, and we use these approximated distributions to optimize a reparameterized variational lower bound. The context given in machine-reading comprehension usually comprises several sentences, leading to performance degradation caused by context length. Therefore, we model a hierarchical structure using sentence encoding, in which as the context becomes longer, the performance degrades. Experimental results show that the proposed VS³-NET model has an exact-match score of 76.8% and an F1 score of 84.5% on the SQuAD test set.

• Journal of Biomedical Engineering Research
• /
• v.42 no.6
• /
• pp.287-294
• /
• 2021

Cardiotoxicity assessment of all drugs has been performed according to the ICH guidelines since 2005. Non-clinical evaluation S7B has focused on the hERG assay, which has a low specificity problem. The comprehensive in vitro proarrhythmia assay (CiPA) project was initiated to correct this problem, which presented a model for classifying the Torsade de pointes (TdP)-induced risk of drugs as biomarkers calculated through an in silico ventricular model. In this study, we propose a TdP-induced risk group classifier of artificial neural network (ANN)-based. The model was trained with 12 drugs and tested with 16 drugs. The ANN model was performed according to nine features, seven features, five features as an individual ANN model input, and the model with the highest performance was selected and compared with the classification performance of the qNet input logistic regression model. When the five features model was used, the results were AUC 0.93 in the high-risk group, AUC 0.73 in the intermediate-risk group, and 0.92 in the low-risk group. The model's performance using qNet was lower than the ANN model in the high-risk group by 17.6% and in the low-risk group by 29.5%. This study was able to express performance in the three risk groups, and it is a model that solved the problem of low specificity, which is the problem of hERG assay.

• Journal of IKEEE
• /
• v.15 no.4
• /
• pp.354-362
• /
• 2011

One of the unsolved problems in Artificial Neural Networks is related to the capacity of a neural network. This paper presents a CoreNet which has a multi-leveled input and a multi-leveled output as a 2-layered artificial neural network. I have suggested an equation for calculating the capacity of the CoreNet, which has a p-leveled input and a q-leveled output, as $a_{p,q}=\frac{1}{2}p(p-1)q^2-\frac{1}{2}(p-2)(3p-1)q+(p-1)(p-2)$. With an odd value of p and an even value of q, (p-1)(p-2)(q-2)/2 needs to be subtracted further from the above equation. The simulation model 1(3)-1(6) has 3 levels of an input and 6 levels of an output with no hidden layer. The simulation result of this model gives, out of 216 possible functions, 80 convergences for the number of implementable function using the cot(x) input leveling method. I have also shown that, from the simulation result, the two diverged functions become implementable by precalculating the weight values. The simulation result and the precalculation of the weight values give the same result as the above equation in the total number of implementable functions.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.26 no.6
• /
• pp.834-841
• /
• 2022

Recently, applying artificial intelligence technologies in various fields of production has drawn an upsurge of research interest due to the increase for smart factory and artificial intelligence technologies. A great deal of effort is being made to introduce artificial intelligence algorithms into the defect detection task. Particularly, detection of defects on the surface of metal has a higher level of research interest compared to other materials (wood, plastics, fibers, etc.). In this paper, we compare and analyze the speed and performance of defect classification by combining machine learning techniques (Support Vector Machine, Softmax Regression, Decision Tree) with dimensionality reduction algorithms (Principal Component Analysis, AutoEncoders) and two convolutional neural networks (proposed method, ResNet). To validate and compare the performance and speed of the algorithms, we have adopted two datasets ((i) public dataset, (ii) actual dataset), and on the basis of the results, the most efficient algorithm is determined.

• Nuclear Engineering and Technology
• /
• v.43 no.4
• /
• pp.343-354
• /
• 2011

It is critically important to identify unexpected loose parts in a nuclear reactor pressure vessel, since they may collide with and cause damage to internal structures. Mass estimation can provide key information regarding the kind as well as the location of loose parts. This study proposes a mass estimation method based on an artificial neural network (ANN), which can overcome several unresolved issues involved in other conventional methods. In the ANN model, input parameters are the discrete cosine transform (DCT) coefficients of the auto-power spectrum density (APSD) of the measured impact acceleration signal. The performance of the proposed method is then evaluated through application to a large-sized plate and a 1/8-scaled mockup of a reactor pressure vessel. The results are compared with those obtained using a conventional method, the frequency ratio (FR) method. It is shown that the proposed method is capable of estimating the impact mass with 30% lower relative error than the FR method, thus improving the estimation performance.

• Nuclear Engineering and Technology
• /
• v.54 no.8
• /
• pp.3043-3050
• /
• 2022

In Korea, all nuclear power plants (NPPs) participate in annual performance tests including in vivo measurements using the FastScan, a stand type whole body counter (WBC), manufactured by Canberra. In 2018, all Korean NPPs satisfied the testing criterion, the root mean square error (RMSE) ≤ 0.25, for the whole body configuration, but three NPPs which participated in an additional lung configuration test in the fission and activation product category did not meet the criterion. Due to the low resolution of the FastScan NaI(Tl) detectors, the conventional peak analysis (PA) method of the FastScan did not show sufficient performance to meet the criterion in the presence of interfering radioisotopes (RIs), ¹³⁴Cs and ¹³⁷Cs. In this study, we developed an artificial neural network (ANN) to improve the performance of the FastScan in the lung configuration. All of the RMSE values derived by the ANN satisfied the criterion, even though the photopeaks of ¹³⁴Cs and ¹³⁷Cs interfered with those of the analytes or the analyte photopeaks were located in a low-energy region below 300 keV. Since the ANN performed better than the PA method, it would be expected to be a promising approach to improve the accuracy and precision of in vivo FastScan measurement for the lung configuration.

• The KSFM Journal of Fluid Machinery
• /
• v.16 no.1
• /
• pp.11-16
• /
• 2013

A numerical study was conducted to improve the performance of an impeller of centrifugal compressor. Nine design variables were chosen with constraints. Only meridional contours and blade profile were adjusted. ANN (Artificial Neural Net) was adopted as a main optimization algorithm with PSO (Particle Swarm Optimization) in order to reduce the optimization time. At first, ANN was learned and trained with the design variable sets which were obtained using DOE (Design of Experiment). This ANN was continuously improved its accuracy for each generation of which population was one hundred. New design variable set in each generation was selected using a non-gradient based method of PSO in order to obtain the global optimized result. After $7^{th}$ generation, the prediction difference of efficiency and pressure ratio between ANN and CFD was less than 0.6%. From more than 1,200 design variable sets, a pareto of efficiency versus pressure ratio was obtained and an optimized result was selected based on the multi-objective function. On this optimized impeller, the efficiency and pressure ratio were improved by 1% and 9.3%, respectively.

• Nuclear Engineering and Technology
• /
• v.53 no.10
• /
• pp.3275-3285
• /
• 2021

A Nuclear Power Plant (NPP) is a complex dynamic system-of-systems with highly nonlinear behaviors. In order to control the plant operation under both normal and abnormal conditions, the different systems in NPPs (e.g., the reactor core components, primary and secondary coolant systems) are usually monitored continuously, resulting in very large amounts of data. This situation makes it possible to integrate relevant qualitative and quantitative knowledge with artificial intelligence techniques to provide faster and more accurate behavior predictions, leading to more rapid decisions, based on actual NPP operation data. Data-driven models (DDM) rely on artificial intelligence to learn autonomously based on patterns in data, and they represent alternatives to physics-based models that typically require significant computational resources and might not fully represent the actual operation conditions of an NPP. In this study, a feed-forward backpropagation artificial neural network (ANN) model was trained to simulate the interaction between the reactor core and the primary and secondary coolant systems in a pressurized water reactor. The transients used for model training included perturbations in reactivity, steam valve coefficient, reactor core inlet temperature, and steam generator inlet temperature. Uncertainties of the plant physical parameters and operating conditions were also incorporated in these transients. Eight training functions were adopted during the training stage to develop the most efficient network. The developed ANN model predictions were subsequently tested successfully considering different new transients. Overall, through prompt prediction of NPP behavior under different transients, the study aims at demonstrating the potential of artificial intelligence to empower rapid emergency response planning and risk mitigation strategies.

• International Journal of Computer Science & Network Security
• /
• v.22 no.10
• /
• pp.406-412
• /
• 2022

Scene classification of very high-resolution (VHR) imagery can attribute semantics to land cover in a variety of domains. Real-world application requirements have not been addressed by conventional techniques for remote sensing image classification. Recent research has demonstrated that deep convolutional neural networks (CNNs) are effective at extracting features due to their strong feature extraction capabilities. In order to improve classification performance, these approaches rely primarily on semantic information. Since the abstract and global semantic information makes it difficult for the network to correctly classify scene images with similar structures and high interclass similarity, it achieves a low classification accuracy. We propose a VHR remote sensing image classification model that uses extracts the global feature from the original VHR image using an EfficientNet-V2L CNN pre-trained to detect similar classes. The image is then classified using a multilayer perceptron (MLP). This method was evaluated using two benchmark remote sensing datasets: the 21-class UC Merced, and the 38-class PatternNet. As compared to other state-of-the-art models, the proposed model significantly improves performance.

• Proceedings of the Korean Society for Agricultural Machinery Conference
• /
• 1993.10a
• /
• pp.1243-1253
• /
• 1993

Visual features of a mushroom(Lentinus Edodes L) are critical in sorting and grading as most agricultural products are. Because of its complex and various visual features, grading and sorting of mushrooms have been done manually by the human expert. Though actions involved in human grading looks simple, a decision making undereath the simple action comes form the results of the complex neural processing of the visual image. And processing details involved in the visual recognition of the human brain has not been fully investigated yet. Recently, however, an artificial neural network has drawn a great attention because of its functional capability as a partial substitute of the human brain. Since most agricultural products are not uniquely defined in its physical properties and do not have a well defined job structure, a research of the neuro-net based human like information processing toward the agricultural product and processing are widely open and promising. In this pape , neuro-net based grading and sorting system was developed for a mushroom . A computer vision system was utilized for extracting and quantifying the qualitative visual features of sampled mushrooms. The extracted visual features and their corresponding grades were used as input/output pairs for training the neural network and the trained results of the network were presented . The computer vision system used is composed of the IBM PC compatible 386DX, ITEX PFG frame grabber, B/W CCD camera , VGA color graphic monitor , and image output RGB monitor.