• The Korean Journal of Applied Statistics
• /
• v.26 no.3
• /
• pp.389-399
• /
• 2013

This article introduces Gaussian process regression and shows its application with time-course microarray gene expression data. Gene screening for yeast cell cycle microarray expression data is accomplished with a ratio of log marginal likelihood that uses Gaussian process regression with a squared exponential covariance kernel function. Gaussian process regression fitting with each gene is done and shown with the nine top ranking genes. With the screened data the Gaussian model-based clustering is done and its silhouette values are calculated for cluster validity.

• Journal of Satellite, Information and Communications
• /
• v.5 no.2
• /
• pp.64-68
• /
• 2010

In this paper, we studied the novel synchronization algorithm for cooperative communication system over fading. We research mainly on the decode-and-forward scheme. Also, we inserted spreading sequence in origin data frame to control efficiently data synchronization. In mobile station, inserted spreading sequence in data frame passed through the corelation process. We had decide the delay value of received data through result of correlation process. In simulation, We applied that channel gain of three node had different value in various fading environment. Finally we will be possible to control the received data synchronization using result of corelation value in each node between relay to mobile station and base station to mobile station. The results of this paper can be applicable to the cooperative systems.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.15 no.3
• /
• pp.137-141
• /
• 2015

Mobile cloud computing uses cloud service to solve the resource constraint problem of mobile devices. Offloading means that a task executed on the mobile device commits to cloud and many studies related to the energy consumption have been researched. In this paper, we designed a response time model considering sensing data synchronization to estimate the efficiency of the offloading scheme in terms of the response time. The proposed model considers synchronization of required sensing data to improve the accuracy of response time estimation when cloud processes the task requested from a mobile device. We found that the response time is effected by new sensing data generation rate and synchronization period through simulation results.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.37 no.4B
• /
• pp.278-287
• /
• 2012

In this paper, we propose an allocation scheme for variable message packings to increase efficiency of military operation using Link-16 which is well-known for tactical data link by delivering imagery information rapidly. We propose a variable message packing scheme using COC waveform to support variable data rate under some coverage limitation. Variety of message packing makes Link-16 vary transmission rate appropriately for tactical environment. We also propose a allocation scheme to assign message packing to time slot properly. Finally we verify the performance and superiority of proposed ideas by simulations.

• Journal of the Korea Society of Computer and Information
• /
• v.7 no.3
• /
• pp.135-142
• /
• 2002

In this paper, we proposed the new structure of the Rake Finger using Walsh Switch, the shared accumulator and the pipeline FWHT algorithm for reducing the signal processing complexity resulting from the increase of the number of data correlators. The number of computational operation in the proposed data correlators is 160 additions when the number of walsh code channels is 4. As a result, it is reduced about 3.2 times other than the number of computational operation of the conventional ones. Also, the result shows that the data processing time of the proposed Rake Finger architecture is 90,496〔ns〕 and the conventional ones is 110,696〔ns〕. It is 18.3% faster than the data processing time of the conventional Rake Finger architecture.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2019.05a
• /
• pp.334-337
• /
• 2019

In this study, we built the data collection system to learn user's habit data by deep learning and to create an indoor environment according to the situation. The system consists of a data collection server and several sensor nodes, which creates the environment according to the data collected. We used Google Inception v3 network to analyze the photographs and hand-designed second DNN (Deep Neural Network) to infer behaviors. As a result of the DNN learning, we gained 98.4% of Testing Accuracy. Through this results, we were be able to prove that DNN is capable of extrapolating the situation.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2019.05a
• /
• pp.224-226
• /
• 2019

Currently, medical data is basically confidential and difficult to access because it is protected under the Medical Protection Act. For the practical education of the students who are in the process of education, the researcher or the faculty members create and upload simulated data (chart of the question bank) close to the actual data. In this paper, the maintenance and repair easier on the basis of node.js and Ajax, mysql, jquery to a web-based research and enables users to easily approach the problem of the chart and the easy to the difficult access to patient contact respectively.

• Journal of the Korea Society for Simulation
• /
• v.18 no.1
• /
• pp.53-62
• /
• 2009

The real data obtained from field exercises has a crucial role in modeling and simulation of a combat or a wargame. This becomes an important input especially in analyzing weapon systems realization. Many existing models have been using the mean value of the time between each fire. The firing data can be incorporated into a known probability distribution or used directly as an empirical distribution. Data of field exercises are very useful instead of the real combat outcomes. This study finds a new modeling approach and techniques to compare the data with the previously generated outcomes. This fundamental research work will continue to consider more of the various weapon systems, the sizes, and other tactical aspects.

• Proceedings of the Korean Information Science Society Conference
• /
• 1999.10b
• /
• pp.467-469
• /
• 1999

Independent component analysis (ICA)는 주어진 데이터를 통계적으로 독립인 요소들의 선형 결합으로 표시하는 통계학적 방법이다. ICA의 주요한 적용분야중의 하나는 source들의 선형 mixture로부터 어떠한 서전 정보도 없는 상태에서 원래의 통계학적 독립변수인 source를 복원하는 blind separation이다. ICA와 source separation을 위한 다양한 신경 학습 알고리듬이 제시되어왔다. ICA의 학습 알고리듬에서는 비선형 함수가 중요한 역할을 한다. 이 논문에서는 generalized 가우시안 prior를 도입하여 다양한 확률분포를 갖는 source들의 mixture를 분리하는 효율적인 source separation 알고리즘을 제시한다. 모의실험을 통하여 제안된 방법의 우수성을 살펴본다.

• Journal of the Korean Data and Information Science Society
• /
• v.5 no.2
• /
• pp.29-44
• /
• 1994

본 연구는 누적위험률함수에 대한 베이지안 추정량을 이용하여 생존함수의 추정량을 제안하고, counting process 이론과 martingale 이론을 이용하여 대표본하에서 제안된 추정량의 일양적 일치성과 점근적 정규성을 밝힌다. 또한 모의실험을 통하여 추정량들의 효율성을 편의와 평균제곱오차의 측면에서 비교하고자 한다.