• Tuberculosis and Respiratory Diseases
• /
• v.86 no.3
• /
• pp.203-215
• /
• 2023

Background: Surgical resection is the standard treatment for early-stage lung cancer. Since postoperative lung function is related to mortality, predicted postoperative lung function is used to determine the treatment modality. The aim of this study was to evaluate the predictive performance of linear regression and machine learning models. Methods: We extracted data from the Clinical Data Warehouse and developed three sets: set I, the linear regression model; set II, machine learning models omitting the missing data: and set III, machine learning models imputing the missing data. Six machine learning models, the least absolute shrinkage and selection operator (LASSO), Ridge regression, ElasticNet, Random Forest, eXtreme gradient boosting (XGBoost), and the light gradient boosting machine (LightGBM) were implemented. The forced expiratory volume in 1 second measured 6 months after surgery was defined as the outcome. Five-fold cross-validation was performed for hyperparameter tuning of the machine learning models. The dataset was split into training and test datasets at a 70:30 ratio. Implementation was done after dataset splitting in set III. Predictive performance was evaluated by R² and mean squared error (MSE) in the three sets. Results: A total of 1,487 patients were included in sets I and III and 896 patients were included in set II. In set I, the R² value was 0.27 and in set II, LightGBM was the best model with the highest R² value of 0.5 and the lowest MSE of 154.95. In set III, LightGBM was the best model with the highest R² value of 0.56 and the lowest MSE of 174.07. Conclusion: The LightGBM model showed the best performance in predicting postoperative lung function.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.12 no.4
• /
• pp.1869-1886
• /
• 2018

This paper proposes a MIDI- and audio-based music genre classification method for Korean traditional music. There are many traditional instruments in Korea, and most of the traditional songs played using the instruments have similar patterns and rhythms. Although music information processing such as music genre classification and audio melody extraction have been studied, most studies have focused on pop, jazz, rock, and other universal genres. There are few studies on Korean traditional music because of the lack of datasets. This paper analyzes raw audio and MIDI phrases in Korean traditional music, performed using Korean traditional musical instruments. The classified samples and MIDI, based on our classification system, will be used to construct a database or to implement our Kontakt-based instrument library. Thus, we can construct a management system for a Korean traditional music library using this classification system. Appropriate feature sets for raw audio and MIDI phrases are proposed and the classification results-based on machine learning algorithms such as support vector machine, multi-layer perception, decision tree, and random forest-are outlined in this paper.

• Wind and Structures
• /
• v.5 no.2_3_4
• /
• pp.337-346
• /
• 2002

Wind tunnel pressure measurements and numerical simulations based on the Reynolds Stress Model (RSM) are compared with full and model scale data in the flow area of impingement, separation and wake for $60^{\circ}$ and $90^{\circ}$ wind azimuth angles. The phase averaged fluctuating pressures simulated by the RSM model are combined with modelling of the small scale, random pressure field to produce the total, instantaneous pressures. Time averaged, rsm and peak pressure coefficients are consequently calculated. This numerical approach predicts slightly better the pressure field on the roof of the TTU (Texas Tech University) building when compared to the wind tunnel experimental results. However, it shows a deviation from both experimental data sets in the impingement and wake regions. The limitations of the RSM model in resolving the intermittent flow field associated with the corner vortex formation are discussed. Also, correlations between the largest roof suctions and the corner vortex "switching phenomena" are observed. It is inferred that the intermittency and short duration of this vortex switching might be related to both the wind tunnel and numerical simulation under-prediction of the peak roof suctions for oblique wind directions.

• Journal of Microbiology and Biotechnology
• /
• v.16 no.5
• /
• pp.724-730
• /
• 2006

A sequence characterized amplified region (SCAR) marker, which was tightly linked with the A1 mating type locus in Phytophthora infestans, was developed. During the random amplified polymorphic DNA-based phylogenic studies of 33 isolates of P infestans collected from year 2002 to 2004, we found an A1 mating type-specific DNA fragment. This 573-bp DNA fragment was generated only in the genomic DNA of the A1 mating types, when OPC-5 primer was used. Based on the specific DNA sequence, we designed the primer sets for generating the A1 mating type-specific 569-bp DNA fragment. When 33 genomic DNAs of P. infestans were subjected to PCR amplification using different primer combinations, the A1 mating type-specific DNA was amplified, when LB-1F and LB-2R primers were used. The specific 569-bp DNA fragment was generated only from all 18 A1 strains, but not from 15 A2 mating type strains. These results corresponded to the mating type discriminating bioassay of 33 isolates of P. infestans. Therefore, the primer combination of LB-1F/LB2R was chosen as a SCAR marker. Overall, this study indicates that the SCAR marker could be developed into a useful tool for mating type determination of P. infestans.

• Nuclear Engineering and Technology
• /
• v.46 no.3
• /
• pp.381-386
• /
• 2014

A highly reliable safety class controller for NPPs (Nuclear Power Plants) is mandatory as even a minor malfunction can lead to disastrous consequences for people, the environment or the facility. In order to enhance the reliability of a safety class digital controller for NPPs, we employed a diversity approach, in which a PLC-type controller and a PLD-type controller are to be operated in parallel. We built and used structured testbenches based on the classes supported by UVM for functional verification of the PLD-type controller designed for NPPs. We incorporated a UVM register model into the testbenches in order to increase the controllability and the observability of the DUT(Device Under Test). With the increased testability, we could easily verify the datapaths between I/O ports and the register sets of the DUT, otherwise we had to perform black box tests for the datapaths, which is very cumbersome and time consuming. We were also able to perform constrained random verification very easily and systematically. From the study, we confirmed the various advantages of using the UVM register model in verification such as scalability, reusability and interoperability, and set some design guidelines for verification of the NPP controllers.

• Nuclear Engineering and Technology
• /
• v.46 no.3
• /
• pp.291-298
• /
• 2014

This paper concerns estimating uncertainties of the core neutronics design parameters of power reactors by direct sampling method (DSM) calculations based on the two-step McCARD/MASTER design system in which McCARD is used to generate the fuel assembly (FA) homogenized few group constants (FGCs) while MASTER is used to conduct the core neutronics design computation. It presents an extended application of the uncertainty propagation analysis method originally designed for uncertainty quantification of the FA FGCs as a way to produce the covariances between the FGCs of any pair of FAs comprising the core, or the covariance matrix of the FA FGCs required for random sampling of the FA FGCs input sets into direct sampling core calculations by MASTER. For illustrative purposes, the uncertainties of core design parameters such as the effective multiplication factor ($k_{eff}$), normalized FA power densities, power peaking factors, etc. for the beginning of life (BOL) core of Yonggwang nuclear unit 4 (YGN4) at the hot zero power and all rods out are estimated by the McCARD/MASTER-based DSM computations. The results are compared with those from the uncertainty propagation analysis method based on the McCARD-predicted sensitivity coefficients of nuclear design parameters and the cross section covariance data.

• Biomedical Science Letters
• /
• v.19 no.3
• /
• pp.213-223
• /
• 2013

The objectives of this study are to develop a molecular diagnosis to differentiate serotypes and mating-types of C. neoformans/C. gattii complex isolates from pigeon droppings in Korea and to elucidate molecular epidemiology of the isolates. Phenotypes and genotypes of C. neoformans/C. gattii complex isolates were identified by biochemical properties and PCR using specific CNLAC1 gene, respectively. To classify serotypes and mating-types of C. neoformans/C. gattii complex isolates, the five reference strains and thirty-three isolates in Korea were investigated by restriction fragment length polymorphism (RFLP) analysis using CNLAC1 gene for varieties, by random amplified polymorphic DNA (RAPD) for serotyping, and by PCR using specific primer sets for mating typing. All isolates in Korea were belonged to C. neoformans var. grubii (serotype A) by RFLP and RAPD patterns which showed high sensitivity and specificity. Therefore, RFLP and RFLP were available to differentiate varieties and serotypes of C. neoformans. Amplification patterns of the five reference strains by specific PCR for mating typing were differentiable, and all isolates were classified into $MAT{\alpha}$. All C. neoformans environmental isolates in Korea were Cr. neoformans serotype A and $MAT{\alpha}$ which is a more virulent pathogen. This study suggests that RFLP and RAPD are rapid and correct molecular diagnosis tools for epidemiology of C. neoformans/C. gattii complex isolates.

• Asian-Australasian Journal of Animal Sciences
• /
• v.13 no.8
• /
• pp.1040-1043
• /
• 2000

The genomic mapping of Red Jungle Fowl (Gallus gallus), local Village Chicken, and broiler was carried out by random amplified polymorphism DNA (RAPD) technique. Two different sets of arbitrary primers were used (Operon OPA01-20 and Genemed GM01-50). All the genomes of the three species of chickens were amplified with OPA01-20 primers. The genomes of the Red Jungle Fowl and local Village Chicken were further amplified with GM01-50 primers. Analysis of the results based on band sharing (BS) and the molecular size of individually amplified DNA fragments showed that Red Jungle Fowl and local Village Chicken shared the species similarity of 66% with Operon primers 01-20, 64% between local Village Chicken and broiler, and 63% when DNA bands between Red Jungle Fowl and broiler were compared. With GM01-50, the BS between Red Jungle Fowl and local village chicken increased to 72%. The results showed that the local village chicken is more closely related to Red Jungle Fowl than to broiler in the genetic distance. On the other hand, broiler is 1% closer in genetic distance to local village chicken than to Red Jungle Fowl. The results also indicated that primers like OPA-7, 8 and 9 can be used as species specific DNA markers for these three species of chickens.

• Journal of Korea Multimedia Society
• /
• v.14 no.8
• /
• pp.965-980
• /
• 2011

In this paper, the use of conditional Fuzzy C-Means (CFCM) aimed at estimation of desired states of an unknown digital communication channel is investigated for blind channel equalization. In the proposed CFCM, a collection of clustered centers is treated as a set of pre-defined desired channel states, and used to extract channel output states. By considering the combinations of the extracted channel output states, all possible sets of desired channel states are constructed. The set of desired states characterized by the maximal value of the Bayesian fitness function is subsequently selected for the next fuzzy clustering epoch. This modification of CFCM makes it possible to search for the optimal desired channel states of an unknown channel. Finally, given the desired channel states, the Bayesian equalizer is implemented to reconstruct transmitted symbols. In a series of simulations, binary signals are generated at random with Gaussian noise, and both linear and nonlinear channels are evaluated. The experimental studies demonstrate that the performance (being expressed in terms of accuracy and speed) of the proposed CFCM is superior to the performance of the existing method exploiting the "conventional" Fuzzy C-Means (FCM).

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.13 no.5
• /
• pp.379-389
• /
• 1988

In this paper, an algorithm for obtaining 2-state switching failure function and a terminal pair reliability evaluation method in a communication network are proposed. The communication network is modeled by a graph. By using the sequence of matrix transformations for the graph, minimal cut-set matrix representing all minimal cut-sets which completely interrupt the communication path is determined and 2-state switching failure function is then obtained from the minimal cut-set matrix. The terminal pari reliability of the communication network is evaluated by corresponding the random variables to 2-state switching failure function. Illustrative examples are provided to demonstrate the algorithm. A computer program evaluating the terminal pair reliability in a complex and large network has also been developed.