• 품질경영학회지
• /
• 제27권1호
• /
• pp.80-90
• /
• 1999

In this paper, reliability estimation using Gibbs sampler is considered for the mixture model with Gamma family, Gibbs sampler is derived to compute the features for the posterior distribution. By simulation study, the maximum likelihood estimator and the Gibbs estimator are obtained. A numerical study with a simulated data is provided.

• Geomechanics and Engineering
• /
• 제9권6호
• /
• pp.743-755
• /
• 2015

It is well known that the quality of sample significantly determines the accuracy of soil parameters for laboratory testing. Although sampling disturbance has been studied over the last few decades, the theoretical investigation of soil disturbance due to sampling penetration has been rarely reported. In this paper, an analytical solution for estimating the soil disturbance due to sampling penetration was presented using cavity expansion method. Analytical results in several cases reveal that the soil at different location along the sample centerline experiences distinct phases of strain during the process of sampling penetration. The magnitude of induced strain is dependent on the position of the soil element within the sampler and the sampler geometry expressed as diameter-thickness ratio D/t and length-diameter ratio L/D. Effects of sampler features on soil disturbance were also studied. It is found that the induced maximum strain decreases exponentially with increasing diameter-thickness ratio, indicating that the sampling disturbance will reduce with increasing diameter or decreasing wall thickness of sampler. It is also found that a large length-diameter ratio does not necessarily reduce the disturbance. An optimal length-diameter ratio is suggested for the further design of improved sampler in this study.

• Journal of applied mathematics & informatics
• /
• 제6권2호
• /
• pp.661-668
• /
• 1999

Reliability estimation using Gibbs sampler considered for modeling mixture exponential reliability problems. Gibbs sampler is developed to compute the features of the posterior distribution. Bayesian estimation of complicated functions requires simpler esti-mation techniques due to the mathematical difficulties involved in the Bayes approach. The Maximum likelihood estimator and the Gibbs estimator of reliability of the system are derived. By simula-tion risk behaviors of derived estimators are compared. model de-termination based on relative error is considered. A numerical study with a simulated data set is provided.

• 한국품질경영학회:학술대회논문집
• /
• 한국품질경영학회 1998년도 The 12th Asia Quality Management Symposium* Total Quality Management for Restoring Competitiveness
• /
• pp.225-232
• /
• 1998

A Markov Chain Monte Carlo method with data augmentation is developed to compute the features of the posterior distribution. This data augmentation approach facilitates the specification of the transitional measure in the Markov Chain. Bayesian analysis of the mixture exponential model discusses using the Gibbs sampler. Parameter and reliability estimators are obtained. A numerical study is provided.

• 한국농공학회지
• /
• 제10권1호
• /
• pp.1377-1387
• /
• 1968

In order to measure runoff and soil losses produced in a small test plot during rainfall, it is usually insufficient to use a tank only, necessitating the combined use of a main tank and a subsidiary tank. Accordingly. exact measurement largely depends on how to connect those two measuring tanks. The main purpose of this thesis is to improve the connecting parts of two measuring tanks so as to assure exact measurement of runoff and soil losses. In this experiment, two types of main tank, i. e. A-type and B-type, were used. A-type is a square tank having a flume at its end. At the flume, ten apertures are provided by using metal columns so as to be able to catch one tenth of total muddy flow discharging at the end of the flume, One tenth of total flow is led to the subsidiary tank through a slot sampler fixed to an aperture. B-type differes in that its flume does not have apertures and slot sampler is fixed directly to the end of the flume, other features being the same as those of A-type. Discharge volumes were measured by using weighing tanks and compared. The effect of baffle screen provided in the flume was also observed in connection with exact measurements. In order to keep main tank and its flume in a horizontal position, bolts and nuts mechanism was used. Vertical and horizontal screens were provided in the main to prevent coarse sands coming into the flume. The conclusion derived through this experiment is as follows: (1) The discharge through slot sampler at each aperture is almost the same for A-type. However, it is slightly more than one tenth of total discharge volume. (2) In case that baffle screen is provided in the flume of A-type tank, the discharge volume of slot sampler is less than that of the same type without screen. (3) For B-type tank, slot sampler discharge increases as slot sampler nears toward the center of flume. (4) When baffle screen is provided in the flume of B-type, slot sampler discharge is less than that of the same type without screen, and this phenomenon is more apparent as compared with A-type. (5) In case that the slot width of slot sampler for B-type is one inch, slot sampler discharge exceeds one tenth of total discharge volume. (6) When the slot width for B-type is 15/16 inch and slot sampler is fixed 3/8 inch apart from either flume wall, slot sampler discharge is approximately equal to one tenth of total discharge volume.

• 지능정보연구
• /
• 제28권4호
• /
• pp.329-346
• /
• 2022

최근 소셜 미디어의 숏폼(Short form) 동영상(인스타그램, 틱톡, 유튜브) 시장이 점차 증가하면서 인공지능 영역에서는 이를 활용한 연구가 활발히 진행되고 있다. 대표적인 연구분야로 동영상 내의 패션 상품을 탐지하고 상품 이미지를 검색하는 Video to shop 을 들 수 있다. 이와 같은 동영상 기반 인공지능 모델에서는 Convolution 연산을 사용하여 상품의 특징을 추출한다. 하지만 연산 자원의 제한으로 인해, 동영상의 모든 프레임을 사용하여 특징을 추출하는 것은 현실적으로 불가능하다. 이로 인해, 기존 연구에서는 전체 프레임 중 일부만 샘플링해서 사용하거나, 주제의 특성을 활용한 샘플링 방법을 개발하여 이를 통해 위 문제점을 개선하고, 모델의 성능도 향상시켰다. 기존의 Video to shop 연구에서는 프레임을 샘플링 할 때, 무작위로 일부분의 프레임을 샘플링하거나 균등한 간격으로 샘플링 한다. 하지만 이러한 샘플링 방법은 상품이 존재하지 않는 노이즈 프레임을 샘플링 하면서 패션 상품 검색 모델의 성능을 저하시킨다. 이에 본 연구는 노이즈 프레임을 제거하고 검색 모델의 성능을 향상시키는 샘플링 방법 MF(Missing Fashion items on frame) sampler를 제안한다. MF sampler는 키 프레임 메커니즘(Mechanism)을 발전시켜 자원 한계의 문제점을 개선했다. 또한, 노이즈 탐지 모델을 활용한 노이즈 프레임 제거를 통해 검색 모델의 성능을 향상시켰다. 이와 같은 결과는 실험을 통해 확인되었고, Video to shop 패션 상품 검색에 있어 성능 향상과 효과적인 학습이 가능하다는 것을 확인할 수 있었다.

• 천문학논총
• /
• 제19권1호
• /
• pp.101-107
• /
• 2004

This paper describes the digital back-end system for getting the data to analyze the user observation mode by digitalize the analog data after receiving the space radio using the radio telescope, The received analog data will be digitalized by high-speed sampler with 1 Gsps for 4 channel frequency band of millimeter wave, and the digital data will be transported through the fiber-optic digital transmission system and WDM(wavelength division multiplex) to observation building, The wideband digital FIR(Finite Impulse Response) filters analyze the data for user observation mode to record the data in high-speed recorder with 1 Gbps. In this paper, we introduce the overall system configuration and features combined by various information and communication technology in radio astronomy briefly, which will be adopted by KVN(Korean VLBI Network).

• 수산해양기술연구
• /
• 제55권1호
• /
• pp.39-49
• /
• 2019

Acoustics are increasingly regarded as a remote-sensing tool that provides the basis for classifying and mapping ocean resources including seabed classification. It has long been understood that details about the character of the seabed (roughness, sediment type, grain-size distribution, porosity, and material density) are embedded in the acoustical echoes from the seabed. This study developed a sophisticated yet easy-to-use technique to discriminate seabed characteristics using a split beam echosounder. Acoustic survey was conducted in Tongyeong waters, South Korea in June 2018, and the verification of acoustic seabed classification was made by the Van Veen grab sampler. The acoustic scattering signals extracted the seabed hardness and roughness components as well as various seabed features. The seabed features were selected using the principal component analysis, and the seabed classification was performed by the K-means clustering. As a result, three seabed types such as sand, mud, and shell were discriminated. This preliminary study presented feasible application of a sounder to classify the seabed substrates. It can be further developed for characterizing marine habitats on a variety of spatial scales and studying the ecological characteristic of fishes near the habitats.

• Ocean Systems Engineering
• /
• 제14권2호
• /
• pp.101-114
• /
• 2024

Seafloor sediment mapping is an essential research topic in shallow coastal waters, especially in port development, benthic habitat mapping, and underwater communications. The seafloor sediments can be interpreted by collecting sediment samples directly in the field using a grab sampler or corer. Another method is optical, especially using underwater cameras and videos. Both methods each have weaknesses in terms of area coverage (mechanic) and accurate positioning (optic). The latest technology used to overcome it is the acoustic method (echosounder) with Global Navigation Satellite System (GNSS) Real Time Kinematic (RTK) positioning. Therefore, in this study will propose the classification of seafloor sediments in coastal waters using acoustic method that is Multibeam Echosounder (MBES) multi-frequency with five frequency (200 kHz, 250 kHz, 300 kHz, 350 kHz, and 400 kHz). In this study, the deep neural network (DNN) used the bathymetric multi frequency, bathymetric difference inters frequencies, and bathymetric features from 5 (five) frequencies as input layer and 4 (four) sediment types in 74 (seventy-four) sample sediment as output layer to make a seafloor sediment map. Results of sediment mapping using the DNN method show an overall accuracy of 71.6% (significant) and a kappa coefficient of 0.59 (moderate). The distribution of seafloor sediment in the study area is mainly silt (41.6%), followed by clayey sand (36.6%), sandy silt (14.2%), and silty sand (7.5%).

• Journal of applied mathematics & informatics
• /
• 제8권1호
• /
• pp.243-252
• /
• 2001

The points of failure of a decomposition process are defined to be the union of the points of failure from two component point processes for software reliability systems. Because sampling from the likelihood function of the decomposition model is difficulty, Gibbs Sampler can be applied in a straightforward manner. A Markov Chain Monte Carlo method with data augmentation is developed to compute the features of the posterior distribution. For model determination, we explored the prequential conditional predictive ordinate criterion that selects the best model with the largest posterior likelihood among models using all possible subsets of the component intensity functions. A numerical example with a simulated data set is given.