• International Journal of Quality Innovation
• /
• 제10권2호
• /
• pp.97-108
• /
• 2009

This paper is to develop a quality measure to evaluate the quality level of child care service in the regional level. By utilizing the biannual intensive child care statistical reports, ten variables are integrated and summarized as a quality measure for child care service in regional level by employing Principal Component Analysis (PCA). Conclusively, it is possible to get a comprehensive measure and the measure obtained from data between 2003 and 2008 illustrates the difference in child care service quality among regions over years. With the measure developed by this research, each region can also get very good insight into what kinds of factors of child care service should be paid more attention to in order to improve the quality of its child care service. Moreover, the measure obtained in this paper is proven reliable and robust in that it reflects the quality of child care service in each region and gives us statistically uniform quality scores with a different data set.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2005년도 제36회 하계학술대회 논문집 D
• /
• pp.2921-2923
• /
• 2005

This paper presents to generate 3D building model using estimation of rooftop surface after 3D line segment extraction using hybrid stereo matching techniques in terms of the co-operation of area-based stereo and feature-based stereo. we first performed a junction extraction from 3D line segment data which was obtained by stereo images, and finally generated building's reliable rooftop surface model using LSE(Least Square Error) method after creating surfaces by grouped and fixed junction points. we generated synthetic images for experimentation by photo-realistic simulation on Avenches data set of Ascona aerial images.

• 한국지능시스템학회논문지
• /
• 제14권3호
• /
• pp.369-374
• /
• 2004

Convex한 클러스터간의 최적의 거리와 Fuzzy Convex Clustering(FCC) 방법에 의한 효과적인 클러스터 결합 알고리즘을 제시하였다. 또한 두 convex한 클러스터간의 거리 측정 방법의 문제점인 정확성과 수행속도 개선하기 위하여 Support Vector Machines(SVM) 을 이용한 빠르고 정확한 거리 측정 방법을 제시하였다. 따라서 데이터의 부적절한 표현 없이 클러스터들의 개수를 크게 더 줄일 수 있었다. 본 논문에서는 제시한 알고리즘의 타당성을 위하여 여러 데이터에 대한 실험결과를 보여주므로서 제시한 알고리즘을 실제 영상 분할에 적용하여 다른 클러스터링 방법의 결과와 비교분석한다.

• 한국정밀공학회:학술대회논문집
• /
• 한국정밀공학회 2004년도 춘계학술대회 논문요약집
• /
• pp.188-188
• /
• 2004

Modal testing and analysis is a primary tool for obtaining reliable models to represent the dynamics of structures. When a structure is tested in order to collect measured data in modal testing, we usually use attached accelerometers to pick up the response data. Change in modal parameters due to the mass of transducers in modal testing is a well-known problem. The disadvantages are the shift of measured modal frequencies and the change of modal shapes, which can cause inaccurate results in further analysis. Modal analysis methods in frequency domain are based on a set of measured frequency response functions(FRF).(omitted)

• 대한전기학회논문지:시스템및제어부문D
• /
• 제54권3호
• /
• pp.174-179
• /
• 2005

This paper presents 3D line segment extraction method, which can be used in generating 3D rooftop model. The core of our method is that 3D line segment is extracted by using line fitting of elevation data on 2D line coordinates of ortho-image. In order to use elevations in line fitting, the elevations should be reliable. To measure the reliability of elevation, in this paper, we employ the concept of self-consistency. We test the effectiveness of the proposed method with a quantitative accuracy analysis using synthetic images generated from Avenches data set of Ascona aerial images. Experimental results indicate that the proposed method shows average 30 line errors of .16 - .30 meters, which are about $10\%$ of the conventional area-based method.

• Communications for Statistical Applications and Methods
• /
• 제28권1호
• /
• pp.81-88
• /
• 2021

k-least quantile of squares (k-LQS) estimates are a generalization of least median of squares (LMS) estimates. They have not been used as much as LMS because their breakdown points become small as k increases. But if the size of outliers is assumed to be fixed LQS estimates yield a good fit to the majority of data and residuals calculated from LQS estimates can be a reliable tool to detect outliers. We propose to use LQS estimates for separating a clean set from the data in the context of outlyingness of the cases. Three procedures are suggested for the identification of outliers using LQS estimates. Examples are provided to illustrate the methods. A Monte Carlo study show that proposed methods are effective.

• International Journal of Computer Science & Network Security
• /
• 제23권12호
• /
• pp.101-106
• /
• 2023

Chronic illnesses are among the most common serious problems affecting human health. Early diagnosis of chronic diseases can assist to avoid or mitigate their consequences, potentially decreasing mortality rates. Using machine learning algorithms to identify risk factors is an exciting strategy. The issue with existing feature selection approaches is that each method provides a distinct set of properties that affect model correctness, and present methods cannot perform well on huge multidimensional datasets. We would like to introduce a novel model that contains a feature selection approach that selects optimal characteristics from big multidimensional data sets to provide reliable predictions of chronic illnesses without sacrificing data uniqueness.[1] To ensure the success of our proposed model, we employed balanced classes by employing hybrid balanced class sampling methods on the original dataset, as well as methods for data pre-processing and data transformation, to provide credible data for the training model. We ran and assessed our model on datasets with binary and multivalued classifications. We have used multiple datasets (Parkinson, arrythmia, breast cancer, kidney, diabetes). Suitable features are selected by using the Hybrid feature model consists of Lassocv, decision tree, random forest, gradient boosting,Adaboost, stochastic gradient descent and done voting of attributes which are common output from these methods.Accuracy of original dataset before applying framework is recorded and evaluated against reduced data set of attributes accuracy. The results are shown separately to provide comparisons. Based on the result analysis, we can conclude that our proposed model produced the highest accuracy on multi valued class datasets than on binary class attributes.[1]

• 한국BIM학회 논문집
• /
• 제10권2호
• /
• pp.21-28
• /
• 2020

As the number of aging bridges increases, more studies are being conducted on developing effective and reliable methods for the assessment and maintenance of bridges. With the advancement in new sensing systems and data learning techniques through AI technology, there is growing interests in how to evaluate bridges using these advanced techniques. This paper presents a CNN(Convolution Neural Network) deep learning based technique for evaluating the damage existence and for estimating the damage location in PSC bridges using static strain data. Simulation studies were conducted to investigate the proposed method with error analysis. Damage was simulated as the reduction in the stiffness of a finite element. A data learning model was constructed by applying the CNN technique as a type of deep learning. The damage status and its location were estimated using data set built through simulation. It was assumed that the strain gauges were installed in a regular interval under the PSC bridge girders. In order to increase the accuracy in evaluating damage, the squared error between the intact and measured strains are computed and applied for training the data model. Considering the damage occurring near the supports, the results of error analysis were compared according to whether strain data near the supports were included.

• 한국지진공학회:학술대회논문집
• /
• 한국지진공학회 1998년도 추계 학술발표회 논문집 Proceedings of EESK Conference-Spring 1998
• /
• pp.421-435
• /
• 1998

An empirical formula for estimating duration magnitude(MD)is determined by analyzing 619 epicentral distance-duration data set, obtained from earthquakes of 1989-1998 recorded at the KMA network. Based on two assumptions: 1) observed signal duration decreases with increasing epicentral distance, and 2) seismographs of KMA are set at low-gain and therefore inclusion of sensitivity correction term in the equation is not necessary, scaling predicted duration at epicenter to Tsuboi's local magnitude yielded the duration magnitude equation: MD =2.0292$\times$log$\tau$＋0.00123Δ-1.4017 for 1/0$\leq$ML$\leq$5.0, where $\tau$is total signal duration(sec)and Δis epicentral distance(km). Event by event comparison of ML values against MD estimates for t152 events shows that for events having a same ML the difference in MD estimates reaches as high as 1.1 magnitude units. So, to test the usefulness of the duration magnitude equation, we have calculated ML-MD relations by which duration magnitude estimates are converted to local magnitudes ("predicted" ML, say) which are then compared with the directly determined local magnitude values. Except for events with stations where duration is anomalously reestimates(predicted ML) which are in an agreement within a 0.2 magnitude units with the corresponding ML values. Although this study could gain some insights into magnitudes of the past events, we still need to re-examine all the observables in order to obtain more reliable and precise information about magnitude and hypocenter location. So we will pursue a new local-magnitude scaling, as well as refinement of the duration magnitude equation, starting soon with re-reading the amplitudes-arrival time records of (and hence relocating) 250+earthquakes of 1979-present recorded at the KMA network. Thus, with more reliable and precise earthquake parameters determined we would better understand the recent seismicity and related tectonic process within and adjacent region to the Korean peninsula.peninsula.

• Journal of Crop Science and Biotechnology
• /
• 제10권2호
• /
• pp.98-105
• /
• 2007

Phylogenetic relationships in Korean watermelons were evaluated by genetic similarity coefficients using 15 SSR(simple sequence repeat), 14 SCAR(sequence characterized amplified region) and 14 CAPS(sequence characterized amplified region) markers. The SSR markers were selected from previously reported melon and watermelon SSRs through testing polymorphisms within a set of commercial $F_1$ varieties. The SCAR and CAPS markers were developed from polymorphic AFLP(amplified fragment length polymorphism) markers between inbred lines 'BN4001' and 'BN4002'. From the AFLP analysis, 105 polymorphic fragments were identified between the inbred lines using 1,440 primer combinations of EcoRI+CNNN and XbaI+ANNN. Based on the sequencing data of these polymorphic fragments, we synthesized sequence specific primer pairs and detected clear and reliable polymorphisms in 27 primer pairs by indels(insertion/deletion) or RFLP(restriction fragment length polymorphism). A total of 43 sequence-based PCR markers were obtained and polymorphic information content(PIC) was analyzed to measure the informativeness of each marker in watermelon varieties. The average PIC value of SCAR markers was 0.41, which was similar to that of SSR markers. Genetic diversity was also estimated by using these markers to assess the phylogenetic relationships among commercial varieties of watermelon. These markers differentiated 26 Korean watermelon varieties into two major phylogenetic groups, but this grouping was not significantly correlated with their morphological and physiological characteristics. The mean genetic similarity was 66% within the complete set of 26 commercial varieties. In addition, these sequence-based PCR markers were reliable and useful to identify cultivars and genotypes of watermelon.