• International Journal of Contents
• /
• v.17 no.4
• /
• pp.1-15
• /
• 2021

Named entity recognition (NER) is a crucial task for NLP, which aims to extract information from texts. To build NER systems, deep learning (DL) models are learned with dictionary features by mapping each word in the dataset to dictionary features and generating a unique index. However, this technique might generate noisy labels, which pose significant challenges for the NER task. In this paper, we proposed DL-dictionary features, and evaluated them on two datasets, including the OntoNotes 5.0 dataset and our new infectious disease outbreak dataset named GFID. We used (1) a Bidirectional Long Short-Term Memory (BiLSTM) character and (2) pre-trained embedding to concatenate with (3) our proposed features, named the Convolutional Neural Network (CNN), BiLSTM, and self-attention dictionaries, respectively. The combined features (1-3) were fed through BiLSTM - Conditional Random Field (CRF) to predict named entity classes as outputs. We compared these outputs with other predictions of the BiLSTM character, pre-trained embedding, and dictionary features from previous research, which used the exact matching and partial matching dictionary technique. The findings showed that the model employing our dictionary features outperformed other models that used existing dictionary features. We also computed the F1 score with the GFID dataset to apply this technique to extract medical or healthcare information.

• Proceedings of the Korean Nuclear Society Conference
• /
• 1996.11a
• /
• pp.242-247
• /
• 1996

Flow oscillations in boiling channels induces a drastic reduction of the (critical heat flux) CHF or premature burnout. However, most of CHF works and correlations have been focused on stable flow conditions without considering flow oscillation. Therefore to improve the understanding on flow oscillation CHF, in this paper a new CHF correction factor to predict the CHF values under flow oscillation conditions has been developed from 126 experimental data. Also to investigate the dominant factor on flow oscillation CHF parametric trends are analyzed by using the developed correction factor. The overall mean accuracy ratio of the developed correction factor is 1.033 with a standard deviation of 0.195. The RMS errors 0.198. Its assessment shows that the predictions agree well with the experimental data within 25% error bounds.

• Journal of the Korean Wood Science and Technology
• /
• v.27 no.3
• /
• pp.99-116
• /
• 1999

Wooden hollow core flush door is one of the main products of furniture manufacturing and woodworking industries. Warping and buckling of the door is serious problems in service. It has been reported that warping is caused by differences of physical and mechanical properties of face and back of skin panel for the door. This study focused on the prediction of warping and buckling phenomena of the flush door using numerical models. Predictions from the models were also compared with the experimental results obtained from the doors with plywood and hardboard skin panels under various environmental conditions. Three elastic constitutive models, so called elastic beam model, plate model and plate-buckling model, were employed to predict warping and buckling of the doors. It was observed that warping was more pronounced in low humidity condition than in high humidity condition. The plate model considering Poisson's effect was reliable to predict warping more closely than elastic beam model in low humidity condition. The plate-buckling model, however, was the best in the fitting of predictions with the experimental results under high humidity condition because buckling was developed in face and back of skin panel at that condition.

• Bulletin of the Korean Chemical Society
• /
• v.26 no.5
• /
• pp.769-775
• /
• 2005

An efficiency of Monte Carlo (MC) docking simulations was examined for the prediction of chiral discrimination by cyclodextrins. Docking simulations were performed with various computational parameters for the chiral discrimination of a series of 17 enantiomers by $\beta$-cyclodextrin ($\beta$-CD) or by 6-amino-6-deoxy-$\beta$-cyclodextrin (am-$\beta$-CD). A total of 30 sets of enantiomeric complexes were tested to find the optimal simulation parameters for accurate predictions. Rigid-body MC docking simulations gave more accurate predictions than flexible docking simulations. The accuracy was also affected by both the simulation temperature and the kind of force field. The prediction rate of chiral preference was improved by as much as 76.7% when rigid-body MC docking simulations were performed at low-temperatures (100 K) with a sugar22 parameter set in the CHARMM force field. Our approach for MC docking simulations suggested that the conformational rigidity of both the host and guest molecule, due to either the low-temperature or rigid-body docking condition, contributed greatly to the prediction of chiral discrimination.

• 제어로봇시스템학회:학술대회논문집
• /
• 2004.08a
• /
• pp.524-529
• /
• 2004

In this paper, We described the method which developed the optimal control system for air separation unit to change production rates frequently and rapidly. Control models of the process were developed from actual plant data using subspace identification method which is developed by many researchers in resent years. The model consist of a series connection of linear dynamic block and static nonlinear block (Wiener model). The model is controlled by model based predictive controller. In MPC the input is calculated by on-line optimization of a performance index based on predictions by the model, subject to possible constraints. To calculate the optimal the performance index, conditions are expressed by LMI(Linear Matrix Inequalities).In order to access at the Bailey DCS system, we applied the OPC server and developed the Client program. The OPC sever is a device which can access Bailey DCS system.The Client program is developed based on the Matlab language for easy calculation,data simulation and data logging. Using this program, we can apply the optimal input to the DCS system at real time.

• Advances in environmental research
• /
• v.6 no.2
• /
• pp.139-146
• /
• 2017

The purpose of this study was to determine the present air pollutant concentrations and predicted levels for next 30 years in urban environment of Karachi city. For that, a total of fifty measurements were made for each twenty selected locations of the city. The locations were selected on the basis of land use pattern such as residential, commercial, industrial settlements, open areas, congested traffic and low traffic areas for investigation of air pollutants variability and intensity. The measurements were taken continuously for six months period using PM Meter, Model AEROCET 531 and Ambient Air Quality Meter, Model AAQ 7545. The concentration of air pollutants were found higher at Al Asif Square and Maripur Road due to higher intensity of traffic and at Korangi Crossing because of industrial areas. The level of pollutants was lower at Sea View owing to lower traffic congestion and transportation of pollutants by sea breezes.

• Journal of Korea Multimedia Society
• /
• v.25 no.10
• /
• pp.1512-1523
• /
• 2022

Artificial intelligence application in digital health care has been increasing with its development of artificial intelligence. The convergence of the healthcare industry and information and communication technology makes the diagnosis of diseases more simple and comprehensible. From the perspective of medical services, its practice as an initial test and a reference indicator may become widely applicable. Therefore, analyzing the factors that are the basis for existing diagnosis protocols also helps suggest directions using artificial intelligence beyond previous regression and statistical analyses. This paper conducts essential diagnostic prediction learning based on the analysis of blood cancer factors reported previously. Blood cancer diagnosis predictions based on artificial intelligence contribute to successfully achieve more than 90% accuracy and validation of blood cancer factors as an alternative auxiliary approach.

• Journal of the Korea Institute of Military Science and Technology
• /
• v.6 no.4
• /
• pp.147-158
• /
• 2003

The prediction of wind field is very important fact in the radioactive and chemical warfare. In spite of advanced numerical weather prediction modelling and computing technology, the high resolution prediction of wind field is limited by the very high integration costs. In this study we coupled the mesoscale numerical model and microscale diagnostic numerical model with minimized integration costs. This coupled model has not only the ability of prediction of high resolution wind field including complex building but also microscale pollutant diffusion fields. For military operation this system can help making a practical and cost-effective decision in a battle field.

• Food Science and Biotechnology
• /
• v.17 no.5
• /
• pp.1052-1059
• /
• 2008

Sequential Sensitivity Analysis (SSA) and conditional stimulus model have been developed to describe sequence effects in difference tests and proposed to generate prediction of differences in sensitivity between various test protocols and to assist the appropriate selection of difference test. Yet, such models did not furnish a complete explanation of the relative sensitivity in 4 different versions of 3-alternative forced choice (AFC) tests where various interstimulus rinses were introduced. In the present study, the vector of the contrasts between various conditional stimuli were measured using same-different and 2-AFC and a new 16-distribution conditional stimulus model was developed by refining Lee and O'Mahony's contrast model. This new model gave superior predictions than previous models.

• Nuclear Engineering and Technology
• /
• v.53 no.11
• /
• pp.3563-3579
• /
• 2021

Measured decay heat data of light water reactor (LWR) spent nuclear fuel (SNF) assemblies are adopted to train machine learning (ML) models. The measured data is available for fuel assemblies irradiated in commercial reactors operated in the United States and Sweden. The data comes from calorimetric measurements of discharged pressurized water reactor (PWR) and boiling water reactor (BWR) fuel assemblies. 91 and 171 measurements of PWR and BWR assembly decay heat data are used, respectively. Due to the small size of the measurement dataset, we propose: (i) to use the method of multiple runs (ii) to generate and use synthetic data, as large dataset which has similar statistical characteristics as the original dataset. Three ML models are developed based on Gaussian process (GP), support vector machines (SVM) and neural networks (NN), with four inputs including the fuel assembly averaged enrichment, assembly averaged burnup, initial heavy metal mass, and cooling time after discharge. The outcomes of this work are (i) development of ML models which predict LWR fuel assembly decay heat from the four inputs (ii) generation and application of synthetic data which improves the performance of the ML models (iii) uncertainty analysis of the ML models and their predictions.