• The Journal of the Acoustical Society of Korea
• /
• v.22 no.3
• /
• pp.194-201
• /
• 2003

In speech recognition, confidence measuring is to decide whether it can be accepted as the recognized results or not. The confidence is measured by integrating frames into phone and word level. In case of word recognition, the confidence measuring verifies the results of recognition and Out-Of-Vocabulary (OOV). Therefore, the post-processing could improve the performance of recognizer without accepting it as a recognition error. In this paper, we measure the confidence modifying log likelihood ratio (LLR) which was the previous confidence measuring. It accumulates only those which the log likelihood ratio is negative when integrating the confidence to phone level from frame level. When comparing the verification performance for the results of word recognizer with the previous method, the FAR (False Acceptance Ratio) is decreased about 3.49% for the OOV and 15.25% for the recognition error when CAR (Correct Acceptance Ratio) is about 90%.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.31 no.2C
• /
• pp.117-122
• /
• 2006

In this paper, we present a simple and general soft bit decision expression for a Gray coded PAM signal over additive white Gaussian noise(AWGN) channel with the log likelihood ratio(LLR). In order to reduce the complexity of the LLR calculation, we make the bitwise LLR expression simple by replacing the mathematical max functions of the conventional Max-Log-MAP expression with simple arithmetic functions associated with some deterministic parameters, such as a received value and distance between symbols on a signal space. Taking the implementation issues, like the area of silicon, the power consumption, the timing latency, and so on, into consideration, we submit that the proposed method is a promising alternative way to conventional methods for reconfigurable systems.

• Clinical and Experimental Pediatrics
• /
• v.48 no.8
• /
• pp.857-864
• /
• 2005

Purpose : The objectives of this study was to evaluate the correlations between the indices of insulin sensitivity using fasting glucose and insulin level, and the body fat mass measured by bioelectrical impedance analysis(BIA) and dual energy X-ray absorptiometry(DEXA), and to determine the clinical usefulness of insulin sensitivity indices when obese children were followed up. Methods : In this study, 28 simple obese children and adolescents were included. Anthropometric data including body weight, height, obesity degree(OD), body mass index(BMI), and waist-to-hip ratio were collected and then body fat mass was measured by using BIA and DEXA. For metabolic data, 12 hour fasting serum glucose, insulin and lipid profiles were measured and indices for insulin sensitivity(G/I ratio, $log_{insulin}$, HOMA-IR, $log_{HOMA-IR}$, QUICKI) were calculated. Results : BMI had a higher correlation with insulin sensitivity indices than OD(G/I ratio, -0.463 vs -0.209; $log_{insulin}$, 0.417 vs 0.196; HOMA-IR, 0.301 vs 0.238; $log_{HOMA-IR}$, 0.403 vs 0.198; QUICKI, -0.451 vs -0.224). But OD had a higher correlation with body fat mass measured by BIA and DEXA than BMI(BIA, 0.612 vs 0.316; DEXA, 0.667 vs 0.512). The G/I ratio was correlated with body fat mass in BIA(r=-0.420, P<0.05) and DEXA(r=-0.512, P<0.01), percentage of body fat(percentage of fat) in BIA(r=-0.366, P<0.05) and DEXA(r=-0.449, P<0.01). HOMA-IR was only correlated with body fat mass in DEXA(r=0.341, P<0.05). Conclusion : This study revealed that G/I ratios had a statistically significant correlation with anthropometric obesity indices(OD and BMI) and also had a correlation with both body fat mass and percentage of fat. These results suggest that G/I ratios could be used as useful index when obese children and adolescence are followed up.

• The Journal of Engineering Geology
• /
• v.8 no.3
• /
• pp.261-273
• /
• 1998

Multiple well log analysis technique consisting of geophysical well log and geological core log has been made to analysis the discontinuities of a test borehole-PABH1 located in Pungam sedimentary basin, Sosok, Hongchon-gun, Kangwon Province. Well log methods consist of normal resistivity log, focussed log, single point resistance log, SP log, gamma log, natural gamma log as well as acoustic televiewer log and borehole television log. Core scanning technique was used as an aid for geological core log. The analysis was made by comparing firstly the televiewer and core discontinuities, and then the results from conventional geophysical log analysis were compared to those from core log and acoustic televiewer log. Fractures deduced from the acoustic televiewer log coincide well with discontinuities shown on the core and conventional geophysical logs. Particularly close coincidence could be observed between fractures derived from acoustic televiewer and conventional geophysical log analysis. It has been noted that the geophysical logs such as, caliper, resistivity, density and high resolution gamma gamma curves are effective in delineating the fractures. For example the ratio between density and resistivity (BRD/SHN) provides also an alternative indicator for discerning the fracture condition in the study area.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.32 no.3C
• /
• pp.216-225
• /
• 2007

This paper presents bit splitting methods to apply multilevel modulation to iterative codes such as turbo code, low density parity check code and turbo product code. Log-likelihood ratio method splits multilevel symbols to bits using the received in-phase and quadrature component based on Gaussian approximation. However it is too complicate to calculate and implement hardware due to exponential and log calculation. therefore this paper presents Euclidean, MAX and Sector method to reduce the high complexity of LLR method. We propose optimal bit splitting method for three iterative codes.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.13 no.1
• /
• pp.101-113
• /
• 2010

Recently actual rainfall pattern is decreasing rainy days and increasing in rainfall intensity and the frequency of flood occurrence is also increased. To consider recent situation, Engineers use deterministic methods like a PMP(Probable Maximum Precipitation). If design storm wouldn't occur, increasing of design criteria is extravagant. In addition, the biggest structure cause trouble with residents and environmental problem. And then it is necessary to study considering probability of rainfall parameter in each sub-basin for design of water structure. In this study, stochastic rainfall patterns are generated by using log-ratio method, Johnson system and multivariate Monte Carlo simulation. Using the stochastic rainfall patterns, hydrological analysis, hydraulic analysis and 2nd flooding analysis were performed based on GIS for their applicability. The results of simulations are similar to the actual damage area so the methodology of this study should be used about making a flood risk map or regidental shunting rout map against the region.

• The Korean Journal of Applied Statistics
• /
• v.28 no.6
• /
• pp.1191-1208
• /
• 2015

Most statistical methods are dependent on the summary statistic. However, with graphical approaches, it is easier to identify the characteristics of the data and detect information that cannot be obtained by the summary statistic. We present various graphical methods to assess the adequacy of models in logistic regression that include checking log-density ratio, structural dimension, marginal model plot, chi-residual plot, and CERES plot. Through simulation data, we investigate and compare the results of graphical approaches under diverse conditions.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2009.05a
• /
• pp.144-148
• /
• 2009

지구온난화로 인한 홍수빈도 및 강우강도의 증가로 홍수피해의 규모는 더욱 증가하고 있다. 그리고 방재대책도 발생한 피해에 대한 복구를 주된 정책으로 하던 소극적 자세에서 벗어나 과학적 공학적 접근을 통해 재해에 대응할 수 있는 기술을 개발하는 적극적 홍수방어 체계를 구축하고 있다. 또한, 설계빈도의 무조건적인 상향조정에 따른 확정론적인 방법에 의존하기 보다는 추계학적 방법을 도입한 수문량 확충 및 매개변수의 불확실성 분석이 필요성이 대두되고 있는 실정이다. 일반적으로 수공구조물의 설계 시에는 설계홍수량을 사용하는데 이는 홍수특성, 홍수발생빈도, 홍수발생 가능성 등을 수공구조물의 규모와 파괴 시 예상되는 피해 정도에 따라 요구되는 안전성을 함께 고려해 수공구조물의 설계기준으로 활용된다. 설계홍수량은 지속시간과 재현기간을 사전에 지정하여 빈도분석을 통해 설계강우량을 산정하고 이를 강우-유출 분석을 함으로써 산정된다. 설계강우량은 빈도분석을 위해 선택된 강우형태에 따른 수문학적 무작위성을 포함하고 있다. 따라서 시간적 변동을 고려한 적절한 강우양상 형태의 선택은 수문학적으로 안전한 수공구조물의 설계 및 평가에 매우 큰 영향을 미친다. 현재의 경우는 강우형태에 대한 선택이 경험에 의해 임의로 이루어지므로, 토목공학자는 여러 가지 발생 가능한 강우형태에 따른 시간적 분포에 대해 고려해야 한다. 본 연구에서는 Huff분위에 따른 무차원 누가 강우곡선을 이용해 강우변동양상 생성을 위한 제약조건을 log-ratio 변환을 이용해 극복하였으며, 결과의 통계 특성치를 Johnson 분포를 통해 표준정규분포로 변환시켰다. 무작위 변수 발생 후 강우양상별 표준정규값의 상관행렬을 이용하여 상관성이 있는 무작위변수로 변환하였다. 얻어진 상관성을 갖는 무작위변수는 log-ratio 역변환을 통해 상관성을 갖는 변수로 재변환해서 무차원화된 강우곡선을 얻을 수 있었다.

• Proceedings of the Korean Geotechical Society Conference
• /
• 2009.03a
• /
• pp.526-532
• /
• 2009

말뚝의 정재하시험을 통하여 항복하중 및 극한하중을 판별하는 다양한 방법이 제안되어 왔다. 말뚝의 지지력은 주면마찰력과 선단지지력의 합으로 나타내어 왔으나 말뚝 재하시험을 통한 항복 하중 및 극한하중의 판별법은 대체로 총 침하량에 대해 판별하거나 재하하중-침하 그래프로부터 산정되는 경우가 대표적이다. 본 연구에서는 현장 대구경 양방향 재하시험 결과를 활용하여 말뚝의 주면부와 선단부로 나누어 항복하중을 판별할 수 있는 방법을 제안하고자 한다.

• Phonetics and Speech Sciences
• /
• v.1 no.1
• /
• pp.55-62
• /
• 2009

This paper proposes an improved method to verify a keyword segment that results from a word spotting system. First a baseline word spotting system is implemented. In order to improve performance of the word spotting systems, we use a two-pass structure which consists of a word spotting system and an utterance verification system. Using the basic likelihood ratio test (LRT) based utterance verification system to verify the keywords, there have been certain problems which lead to performance degradation. So, we propose a method which uses phone-level log-likelihood ratios (PLLR) patterns in computing confidence measures for each keyword. The proposed method generates weights according to the PLLR patterns and assigns different weights to each phone in the process of generating confidence measures for the keywords. This proposed method has shown to be more appropriate to word spotting systems and we can achieve improvement in final word spotting accuracy.