• 대한임베디드공학회논문지
• /
• 제9권4호
• /
• pp.245-251
• /
• 2014

Wireless Sensor Networks(WSNs) are applied to many monitoring applications. Present sensor nodes can perform many functions at the same time and contain complex software. During the lifetime of sensor nodes, they are required to reprogram their software because of their new functions, software, software bug fixes. The nodes are inaccessible physically or it is very difficult to upgrade their software by one by one. To upgrade the software of sensor nodes in WSNs remotely, this paper presents an energy efficient method by selecting an optimal relay node. The CHR(Cluster Head Relay) method is compared with SPIN and RANDOM method. Three methods are simulated in NS-2 with the same environmental parameters. Simulation results show that CHR shows faster update time and less power consumption compared with other two methods.

• Journal of information and communication convergence engineering
• /
• 제19권3호
• /
• pp.136-141
• /
• 2021

Research on wireless sensor networks has focused on the monitoring and characterization of large-scale physical environments and the tracking of various environmental or physical conditions, such as temperature, pressure, and wind speed. We propose a stochastic mobility model that can be applied to a MANET (Mobile Ad-hoc NETwork). environment, and apply this mobility model to a newly proposed clustering-based routing protocol. To verify its stability and durability, we compared the proposed stochastic mobility model with a random model in terms of energy efficiency. The FND (First Node Dead) was measured and compared to verify the performance of the newly designed protocol. In this paper, we describe the proposed mobility model, quantify the changes to the mobile environment, and detail the selection of cluster heads and clusters formed using a fuzzy inference system. After the clusters are configured, the collected data are sent to a base station. Studies on clustering-based routing protocols and stochastic mobility models for MANET applications have shown that these strategies improve the energy efficiency of a network.

• 한국통신학회논문지
• /
• 제38A권1호
• /
• pp.1-9
• /
• 2013

본 논문에서는 CS-LBP (Center-Symmetric Local Binary Pattern) 특징과 공간 피라미드를 이용한 BoF (Bag of Features)를 생성하고 이를 랜덤 포레스트(Random Forest) 분류기에 적용하여 인간의 행동을 인식하는 알고리즘을 제안한다. BoF를 생성하기 위해 영상을 균일한 패치로 나누고, 각 패치 마다 CS-LBP 특징을 추출한다. 행동 분류 성능을 향상시키기 위해 패치들마다 추출한 특징벡터들에 대해 K-mean 클러스터링을 적용하여 코드 북을 생성한다. 본 논문에서는 영상의 지역적인 특성을 고려하기 위해 공간 피라미드 방법을 적용하고 각 공간 레벨에서 추출된 BoF에 대해 가중치를 적용하여 최종적으로 하나의 특징 벡터로 결합한다. 행동 분류를 위해 결정트리의 앙상블로 이루어진 랜덤 포레스트는 학습 단계에서 각 행동 클래스를 위한 분류 모델을 만든다. 가중 BoF가 적용된 랜덤 포레스트는 다양한 인간 행동 영상을 포함하고 있는 Standford Actions 40 데이터를 성공적으로 분류하였다. 또한 기존 방법에 비해 분류 성능이 유사하거나 우수하며, 한 장의 영상에 대해 빠른 인식속도를 보였다.

• 한국환경농학회지
• /
• 제34권3호
• /
• pp.168-177
• /
• 2015

BACKGROUND: In recent year, there has been great concern regarding agricultural land uses and their importance for the conservation of biodiversity. Rice fields are managed unique wetland for wildlife, especially waterbirds. A comprehensive monitoring of the waterbird assemblage to understand patterning changes was attempted for rice ecosystem in South Korea. This rice ecosystem has been recognized as one of the most important for waterbirds conservation. METHODS AND RESULTS: Biweekly monitoring was implemented for the 4 years from April 2009 to March 2010, from April 2011 to March 2014. 32 species of waterbirds were observed. Self-organizing map (SOM) and random forest were applied to the waterbirds dataset to identify the characteristics in waterbirds distribution. SOM and random forest analysis clearly classified into four clusters and extract ecological information from waterbird dataset. Waterbird assemblages represented strong seasonality and habitat use according to waterbird group such as shorebirds, herons and waterfowl. CONCLUSION: Our results showed that the combination of SOM and random forest analysis could be useful for ecosystem assessment and management. Furthermore, we strongly suggested that a strict management strategy for the rice fields to conserve the waterbirds. The strategy could be seasonally and species specific.

• The Plant Pathology Journal
• /
• 제23권2호
• /
• pp.45-50
• /
• 2007

Genetic diversity among selected isolates of Macrophomina phaseolina, a causal agent of charcoal rot (stalk rot) disease in sorghum was studied using PCR-RAPD markers. A set of ten isolates, from ten different rabi sorghum genotypes representing two traditional sorghum growing situations viz., Dharwad- a transitional high rainfall region and Bijapur- a semi-arid low rainfall region in South India. From a set of 40 random primers tested, amplicon profiles of 15 were reproducible. A total of 149 amplicon levels, with an average of 9.9 bands per primer, were available for analysis, of which 148 were polymorphic (99.3%). It was possible to discriminate all the isolates with any of the 15 primers employed. UPGMA clustering of data indicated that the isolates shared varied levels of genetic similarity within a range of 0.14 to 0.72 similarity coefficient index and it was suggestive that grouping of isolates was not related to sampling location in anyway. A high level of genetic heterogeneity of 0.28 was recorded among the isolates.

• Communications for Statistical Applications and Methods
• /
• 제20권4호
• /
• pp.301-309
• /
• 2013

This paper predicts gross box office revenue for domestic films using the Korean film data from 2008-2011. We use three regression methods, Linear Regression, Random Forest and Gradient Boosting to predict the gross box office revenue. We only consider domestic films with a revenue size of at least KRW 500 million; relevant explanatory variables are chosen by data visualization and variable selection techniques. The key idea of analyzing this data is to construct the meaningful explanatory variables from the data sources available to the public. Some variables must be categorized to conduct more effective analysis and clustering methods are applied to achieve this task. We choose the best model based on performance in the test set and important explanatory variables are discussed.

• Communications for Statistical Applications and Methods
• /
• 제16권1호
• /
• pp.185-193
• /
• 2009

Machine learning methods such as support vector machines and random forests yield nonparametric prediction functions of the form y = $f(x_1,{\ldots},x_p)$. As a sequel to the previous article (Huh and Lee, 2008) for visualizing nonparametric functions, I propose more sensible graphs for visualizing y = $f(x_1,{\ldots},x_p)$ herein which has two clear advantages over the previous simple graphs. New graphs will show a small number of prototype curves of $f(x_1,{\ldots},x_{j-1},x_j,x_{j+1}{\ldots},x_p)$, revealing statistically plausible portion over the interval of $x_j$ which changes with ($x_1,{\ldots},x_{j-1},x_{j+1},{\ldots},x_p$). To complement the visual display, matching importance measures for each of p predictor variables are produced. The proposed graphs and importance measures are validated in simulated settings and demonstrated for an environmental study.

• 산업경영시스템학회지
• /
• 제23권60호
• /
• pp.73-81
• /
• 2000

The testing processes for VLSI finished devices are considerably complex because they require different types of ATE to be linked together. Due to the interaction effect between two or more linked ATEs, it is difficult to trace down the cause of the unexpected longer ATE setup time and random yields, which frequently occur in the VLSI circuit-testing laboratory. The goal of this paper is to develop and demonstrate the methodology designed to eliminate the possible interaction factors that might affect the random yields and/or unexpected longer setup time as well as increase the productivity. The statistical method such as design of experiment or multivariate analysis cannot be applied to the final testing floor here directly due to the environmental constraints. Expanded product data information (PDI) is constructed by combining product data information and ATE control information. An architecture utilizing expanded PDI is designed, which enables the engineer to conduct statistical approach investigation and reduce the setup time, as well as increase yield.

• 한국산림과학회지
• /
• 제99권3호
• /
• pp.259-266
• /
• 2010

우리나라에서 산불 발생 공간분포는 인간 활동과 큰 관련성이 있기 때문에, 지역별 군집형태의 강한 공간의존성을 갖는다. 본 연구에서는 공간의존성의 개념에 입각하여 점자료 분석법을 통한 산불발생의 공간분포패턴을 분석하였다. Ripley의 K 함수를 이용하여 산불 발생 원인별 공간분포 형태를 파악하였으며, Kernel 함수를 통해 산불발생의 공간적 집중도를 분석하였다. 그 결과 정도는 상이하지만 모든 원인의 산불이 임의(random) 분포가 아닌 군집화(clustered)되어 발생하는 특징이 있는 것으로 나타났다. 또한, 산불 발생의 군집성을 원인별로 크게 두 집단으로 나눌 수 있었다. 첫째는 전국적 발생 패턴을 가지는 원인으로 입산자 실화, 논밭두렁 소각과 같은 활동과 관련된 것이고 또 다른 하나는 국지적 군집성을 가지는 원인으로 담뱃불이나 어린이 불장난, 방화이다. 그 군집성의 범위는 30 km내외로 나타났으며, 그 범위 밖에서는 임의 분포하고 있었다. Kernel 함수에 의한 원인별 집중도 분석에서는 강한 군집도를 나타냈던 3가지 원인(담뱃불, 어린이 불장난, 방화)의 경우 대부분 인구밀도가 높은 수도권을 중심으로 발생하는 것을 확인할 수 있었다.

• 응용통계연구
• /
• 제17권2호
• /
• pp.337-346
• /
• 2004

랜덤표본에 관한 이원분할표의 독립성검정에는 통상 피어슨의 카이제곱적합도검정과 우도비검정을 사용한다. 그러나 랜덤표본이 아닌 집락자료에 관한 분할표의 경우에는 이들 검정법은 잘못된 결과를 나타낸다. 이러한 경우에는 공변량의 고정효과 외에 집락에 따른 변량효과를 함께 포함하는 일반화선형혼합모형을 고려함으로써 집락간의 이질성과 집락내의 종속성을 반영할 수 있다. 본 연구에서는 집락자료의 분할표에 대한 일반화선형혼합모형을 소개하고 실례를 통하여 이들 모형의 적합에 대해 논의한다.