• The Korean Journal of Applied Statistics
• /
• v.34 no.2
• /
• pp.225-238
• /
• 2021

The ROC, TOC, and TROC curves, which are visually descriptive methods of exploring the performance of the binary classification model, are implemented with TP, TN, FP, FN which consist of the confusion matrix, as well as their ratios TPR, TNR, FPR, FNR. In this study, we consider two types odds and then propose an odds curve representing these odds. And show the relationship between the odds curve and ROC curve. Based on the odds curve, we propose not only two statistics that measure the discriminant power of the odds curve but also the criteria for validation ratings of the odds curve. According to the shape of the odds curves, two classification distributions can be estimated and a criterion for validation ratings can be determined. The odds curve can be meaningfully used like other visual methods, and two kinds of measures for the discriminant power can be also applied together as an alternative criterion.

• The Journal of the Korea Contents Association
• /
• v.16 no.6
• /
• pp.225-235
• /
• 2016

This study aims at developing Reminiscence Function Scale-Short Form(RFS-S) and verifying it as a supplemental diagnosis instrument to judge differences between normal and mild cognitive impairment and useful measure to rate effects of reminiscence therapy in the clinical field through internal consistency reliability and criterion validity. The study also aims at figuring out factors affection reminiscence function by figuring out demographic features of RFS-S. The research subjects were the control group of 60 normal people and experimental group of 58 people with mild cognitive impairment and both groups were over 65 years old. For internal consistency reliability(Cronbach's alpha) between questions was reliable with 0.63. Convergent validities of RFS-S with SRT and CDR-SB were 0.20(p<.05) and -0.25(p<.001), respectively, as both showed significant correlations. As a result of criterion validity to analyze AUROC, it was 0.68(p<.001) and less accurate. Its optimal cut-off points were 17 and sensitivity according to them was 0.59, and specificity was 0.72, respectively. However, reminiscence function according to demographic variables did not show any significant differences. Therefore, it is expected that RFS-S will be used as a supplemental diagnosis instrument with higher reliability and validity for discerning differences between normal and mild cognitive impairment and as a useful one to verify effects of reminiscence therapy.

• Korean Journal of Cognitive Science
• /
• v.26 no.3
• /
• pp.357-375
• /
• 2015

In our daily life, the probability of target prevalence in visual search varies from very low to high. However, most laboratory studies of visual search used a fixed probability of target prevalence at 50%. The present study examined the properties of information processing during visual search where the probability of target prevalence was manipulated to vary from low (20%), medium (50%), to high (80%). The search items were made of simple shape stimuli, and search accuracy, signal detection measures, and reaction times (RTs) were analyzed for characterizing the effect of target prevalence on the information processing strategies for visual search. The analyses showed that the rates of misses increased whereas those of false alarms decreased in the search condition of low target prevalence, whereas the pattern was reversed in the high prevalence condition. Signal detection measures revealed that the target prevalence shifted response criterion (c) without affecting sensitivity (d'). In addition, RTs for correct rejection responses in the target-absent trials became delayed as the prevalence increased, whereas those for hits in the target-present trials were relatively constant regardless of the prevalence. The RT delay in the target-absent trials indicates that increased target prevalence made the 'quitting threshold' for search termination more conservative. These results support an account that the target prevalence effect in visual search arises from a shift of decision criteria and the subsequent changes in search information processing, while rejecting the account of a speed-accuracy tradeoff.

• Tuberculosis and Respiratory Diseases
• /
• v.46 no.4
• /
• pp.542-547
• /
• 1999

Background: Decision in mediastinal lymph node involvement of lung cancer by CT scan is very important and valuable for the treatment planning and prognosis prediction. In general, long diameter of mediastinal lymph node more than 15mm is used as criterion of lung cancer involvement. Adenocarci-noma has a tendency of early distant metastasis and micrometastasis, so adenocarcinoma may involve lymph node earlier and cannot be detected before lymph nodes are enlarged enough. The authors tried to determine the difference between two size criteria(15mm, 10mm) in adenocarcinoma for the detection of cancer involvement. Methods: Numbers of sample are 60 cases(male 46, female 14, median age: 61.5 years). According to pathology, squamous cancer 41, large cell cancer 2, adenocarcinoma 17. According to TNM stage, I 23, III 24, IIIA 13. Results : Mean long diameter of lymph node involvement is 16.0($\pm8.0$) mm in non-adenocarcinoma group, and that of adenocarcinoma group is 12.0($\pm3.2$) mm(p<0.05). If long diameter of lymph node larger than 15mm as involvement criterion is applied, sensitivity, specificity, positive predictive index, negative predictive index, accuracy of nonadenocarcinoma group are 54%, 100%, 100%, 83%, 86%, and those of adenocarcinoma group are 43%, 90%, 75%, 69%, 71%. If long diameter of lymph node larger than 10mm as involvement criterion is applied, sensitivity, specificity, positive predictive index. negative predictive index. accuracy of nonadenocarcinoma group are 65%, 77%, 61%, 92%, 79%, and those of adenocarcinoma group are 100%, 80%, 78%, 100%, 88%. Conclusion: Long diameter of lymph node larger than 10mm is more valuable criterion as lymph node involvement in adenocarcinoma of lungs.

• Asian-Australasian Journal of Animal Sciences
• /
• v.28 no.12
• /
• pp.1713-1720
• /
• 2015

The present work describes experiments undertaken to evaluate the usefulness of selected physicochemical indices of semen, cell membrane integrity and sperm chromatin structure for the assessment of boar semen sensitivity to processes connected with pre-insemination procedures. The experiments were carried out on 30 boars: including 15 regarded as providers of sensitive semen and 15 regarded as providers of semen that is little sensitive to laboratory processing. The selection of boars for both groups was based on sperm morphology analyses, assuming secondary morphological change incidence in spermatozoa as the criterion. Two ejaculates were manually collected from each boar at an interval of 3 to 4 months. The following analyses were carried out for each ejaculate: sperm motility assessment, sperm pH measurement, sperm morphology assessment, sperm chromatin structure evaluation and cell membrane integrity assessment. The analyses were performed three times. Semen storage did not cause an increase in the incidence of secondary morphological changes in the group of boars considered to provide sperm of low sensitivity. On the other hand, with continued storage there was a marked increase in the incidence of spermatozoa with secondary morphological changes in the group of boars regarded as producing more sensitive semen. Ejaculates of group I boars evaluated directly after collection had an approximately 6% smaller share of spermatozoa with undamaged cell membranes than the ejaculates of boars in group II ($p{\leq}0.05$). In the process of time the percentage of spermatozoa with undamaged cell membranes decreased. The sperm of group I boars was characterised with a lower sperm motility than the semen of group II boars. After 1 hour of storing diluted semen, the sperm motility of boars producing highly sensitive semen was already 4% lower ($p{\leq}0.05$), and after 24 hours of storage it was 6.33% lower than that of the boars that produced semen with a low sensitivity. Factors that confirm the accuracy of insemination male selection can include a low rate of sperm motility decrease during the storage of diluted semen, low and contained incidence of secondary morphological changes in spermatozoa during semen storage and a high frequency of spermatozoa with undamaged cell membranes.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.19 no.2
• /
• pp.381-389
• /
• 1994

In the communication network, multimedia service such as high quality voice, high speed data, image etc. will be added to the existing service. This service generates new requirements for the communication networks. The priority control mechanism can be used to control multimedia traffics generated by many communication systems. The priority mechanism which assigns prioirities to generated cells according to service quality is one of the traffic control. The priority assignment can be divided by priority criterion for each traffic characteristics such as loss sensitivity and delay sensitivity. In this paper, we alnalyzed the partial buffur sharing (PBS) mechani느 as a traffic control reducing the cell loss, and proposed analysis method. We analyzed the PBS mechanism using classical approach as a Markov chain. In order to validata proposed analysis method, simulation is performed using simulation package SIMSCRIPT 11.5. From this results, we confirmed that proposed analysis method can be verified. Also, we presented cell loss probability of ATM network when this results are to be applied to ATM networks.

• Clinical and Experimental Pediatrics
• /
• v.61 no.3
• /
• pp.84-89
• /
• 2018

Purpose: Timely antibiotic therapy in selected cases of diarrhea associated with bacterial infections can reduce the duration and severity of illness and prevent complications. The availability of a predictive index before identification of causative bacteria would aid in the choice of a therapeutic agent. Methods: The study included patients admitted to the pediatrics unit at Konyang University Hospital for acute inflammatory diarrhea from August 1, 2015 to July 31, 2016 who underwent multiplex polymerase chain reaction testing. Of 248 patients, 83 had positive results. The clinical symptoms and blood test results were examined in 61 patients with Campylobacter spp. (25 patients), Salmonella spp. (18 patients), and Clostridium perfringens (18 patients) infections. The mean age of the 61 patients (male:femal=31:30) was $84.0{\pm}54.8months$, and the mean hospital stay was $4.6{\pm}1.7days$. Results: There were no statistical differences in sex, age, clinical symptoms, or signs. Patients with Campylobacter infection were significantly older (P=0.00). C-reactive protein (CRP) levels in patients with Campylobacter infection were higher than those in the other 2 groups, at $9.6{\pm}6.1mg/dL$. The results of receiver-operating characteristic curve analysis showed that the cutoff age was ${\geq}103.5months$ (sensitivity, 72%; specificity, 86%) and the CRP cutoff level was ${\geq}4.55mg/dL$ (sensitivity, 80%; specificity, 69%). Conclusion: Age (${\geq}103.5months$) and higher CRP level (${\geq}4.55mg/dL$) were good predictors of Campylobacter enterocolitis. If neither criterion was met, Campylobacter enterocolitis was unlikely (negative predictive value 97.2%). When both criteria were met, Campylobacter enterocolitis was highly likely.

• Structural Engineering and Mechanics
• /
• v.32 no.2
• /
• pp.283-300
• /
• 2009

The characteristic response of a structure to blast load may be divided into two distinctive phases, namely the direct blast response during which the shock wave effect and localized damage take place, and the post-blast phase whereby progressive collapse may occur. A reliable post-blast analysis depends on a sound understanding of the direct blast effect. Because of the complex loading environment and the stress wave effects, the analysis on the direct effect often necessitates a high fidelity numerical model with coupled fluid (air) and solid subdomains. In such a modelling framework, an appropriate representation of the blast load and the high nonlinearity of the material response is a key to a reliable outcome. This paper presents a series of calibration study on these two important modelling considerations in a coupled Eulerian-Lagrangian framework using a hydrocode. The calibration of the simulated blast load is carried out for both free air and internal explosions. The simulation of the extreme dynamic response of concrete components is achieved using an advanced concrete damage model in conjunction with an element erosion scheme. Validation simulations are conducted for two representative scenarios; one involves a concrete slab under internal blast, and the other with a RC column under air blast, with a particular focus on the simulation sensitivity to the mesh size and the erosion criterion.

• Genomics & Informatics
• /
• v.16 no.4
• /
• pp.31.1-31.7
• /
• 2018

The prevalence of metabolic syndrome (MS) in the nonobese population is not low. However, the identification and risk mitigation of MS are not easy in this population. We aimed to develop an MS prediction model using genetic and clinical factors of nonobese Koreans through machine learning methods. A prediction model for MS was designed for a nonobese population using clinical and genetic polymorphism information with five machine learning algorithms, including naïve Bayes classification (NB). The analysis was performed in two stages (training and test sets). Model A was designed with only clinical information (age, sex, body mass index, smoking status, alcohol consumption status, and exercise status), and for model B, genetic information (for 10 polymorphisms) was added to model A. Of the 7,502 nonobese participants, 647 (8.6%) had MS. In the test set analysis, for the maximum sensitivity criterion, NB showed the highest sensitivity: 0.38 for model A and 0.42 for model B. The specificity of NB was 0.79 for model A and 0.80 for model B. In a comparison of the performances of models A and B by NB, model B (area under the receiver operating characteristic curve [AUC] = 0.69, clinical and genetic information input) showed better performance than model A (AUC = 0.65, clinical information only input). We designed a prediction model for MS in a nonobese population using clinical and genetic information. With this model, we might convince nonobese MS individuals to undergo health checks and adopt behaviors associated with a preventive lifestyle.

• Journal of the Architectural Institute of Korea Structure & Construction
• /
• v.34 no.2
• /
• pp.19-28
• /
• 2018

The progressive collapse resistance performance of a steel structure constructed using the moment frame with the WUF-B connection and the composite slabs was evaluated. GSA 2003 was adapted for the evaluation. Additionally the structural robustness and the sensitivity against the progressive collapse were analyzed. In the numerical analysis, a reduced model comprised of the beam and spring elements for WUF-B connection was adapted. The composite slab was modeled using the composite-shell element. Instead of the time-consuming dynamic analysis for the effect of the sudden column removal, the equivalent energy-based static analysis was effectively applied. The analysis results showed that the structure was the most vulnerable to in the case of the internal column removal, however it satisfied the chord rotation criterion of GSA 2003 due to the contribution of the composite slab which improved the stiffness of structure. In the robustness evaluation, the structural performance showed more than 2.5 times of the requirement according to GSA 2003, and the structural sensitivity analysis indicated the decrease of 33% of the initial structural performance.