In this study, we analyzed the relationship between phytoplankton community and physicochemical factors in 12 lakes located in Jeollanam-do based on the data surveyed from March to November 2014. Totally, 297 species of phytoplankton were identified including 98 Bacillariophyceae, 148 Chlorophyceae, 23 Cyanophyceae and 28 other phytoplankton taxa. The standing crops ranged from 124 to 59,148 cells mL-1 and showed the highest in August with the increase of Cyanophycean cells. The self-organizing map (SOM) was optimized into 9×6 grid and was classified into 5 clusters based on the similarity of environmental factors and phytoplankton indices. The SOM results showed that phytoplankton communities had positive relationship with water temperature, SS, DO, BOD, TP and Chl-a, whereas low relationship with pH, TN, NH3-N, NO3-N, PO4-P and Conductivity. In Pearson’s correlation coefficient, relationship between environmental factors and phytoplankton communities showed similar results with SOM.

This study examined the effects of environmental factors on the abundance of black pine bast scale (BPBS), Matsucoccus thunbergianae Miller and Park, in coastal disaster prevention forest stands composed mostly of Japanese black pine. Geographical factors, soil conditions and forest stand conditions were measured to evaluate the hazard rating for the occurrence of BPBS from 35 plots in the coastal forest stands. To assess the hazard rating, a combination of a self-organizing map (SOM), which classified the samples according to their characteristics, and a random forest model, which predicted the probability of the occurrence of BPBS from SOM results, was used in this study. Our results showed that major factors determining the abundance of BPBS were climate, tree size, and tree health. BPBS was more common in low latitude coastal forests, suggesting that warmer conditions were favorable to BPBS population buildup. Tree size also influenced the abundance of BPBS, which was higher in forests composed of larger trees (greater DBH). Finally, BPBS was also more abundant in areas with high soil salinity and clay-loam soil, and north-facing slopes where tree vigor was lower.

본 논문에서는 신경망 SOM학습을 이용하여 피험자의 각성수준을 높은각성과 낮은각성으로 자동인식하는 것을 제안한다. 각성수준의 자동인식 단계는 세 단계로 구성된다 첫 번째는 ECG 측정 및 분석단계로 슈팅게임을 플레이하는 피험자를 ECG로 측정하고, SOM 학습을 하기 위해 특징을 추출한다. 두 번째는 SOM 학습 단계로 특징이 추출된 입력벡터들을 학습한다. 마지막으로 각성인식 단계는 SOM 학습이 완료된 후에 새로운 입력벡터가 들어왔을 때, 피험자의 각성수준을 인식한다. 실험결과는 각성수준의 SOM 학습결과와 새로운 입력벡터가 들어왔을 때 각성수준의 인식결과, 그리고 각성수준을 수치와 그래프로 보여준다. 마지막으로 SOM의 평가는 기존연구의 감성평가 결과와 SOM의 자동인식 결과를 순차적으로 비교하여 평균 86%로 분석되었다. 본 연구를 통해서 SOM을 이용하여 피험자마다 다른 각성수준을 자동인식 할 수 있었다.

Data mining is widely used for turning huge amounts of data into useful information and knowledge in the information industry in recent years. When analyzing data set with continuous values in order to gain knowledge utilizing data mining, we often underg

Data mining is widely used for turning huge amounts of data into useful information and knowledge in the information industry in recent years. When analyzing data set with continuous values in order to gain knowledge utilizing data mining, we often undergo a process called discretization, which divides the attribute’s value into intervals. Such intervals from new values for the attribute allow to reduce the size of the data set. In addition, discretization based on rough set theory has the advantage of being easily applied. In this paper, we suggest a discretization algorithm based on Rough set and SOM(Self-Organizing Map) as a means of extracting valuable information from large data set, which can be employed even in the case where there lacks of professional knowledge for the field.

인간의 적응적인 마음상태는 긍정적/부정적 사고의 상호작용에서 자기 고양적인 방향으로 편향된 비대칭성을 가정한다. 본 논문에서는 긍정적/부정적 인지, 정서의 균형에 대한 심리 수학적 모형인 SOM 모형을 감성판단 영역에 적용하여 보았다. ,iOM 모형은 개인이 긍정적, 부정적 감성의 이분법적 판단에 있어서 평균적으로 6B의 비율을 유지한다고 제안하며, 이러한 비율은 고대수학의 .6)0 ,.182의 황금분할비와 일치하는 것이다. 32명의 일반성인을 대상으로 한 분석결과, 긍정적/부정적 감성단어에 대한 자기 관련성 평정의 SOM 비율은 .62(SD=.On)로 나타났으며, 우연회상 과제수행의 SOM 비율은 .18(SD=.34)로 이론적 기준과 유사하게 나타났다. 또한 상관분석에서 개인의 10M 비율은 긍정적/부정적 감성단어에 대한 평정과도 유의미한 상관관계를 지니고 있는 것으로 나타났다. 이러한 결과를 바탕으로 인지적 균형비의 의미와 감성과학 분야의 적용 가능성에 대해 제안하였다.

Ministry of Environment has been operating water quality monitoring network in order to obtain the basic data for the water environment policies and comprehensively understand the water quality status of public water bodies such as rivers and lakes. The observed water quality data is very important to analyze by applying statistical methods because there are seasonal fluctuations. Typically, monthly water quality data has to analyze that the transition comprise a periodicity since the change has the periodicity according to the change of seasons. In this study, trends, SOM and RDA analysis were performed at the Mulgeum station using water quality data for temperature, BOD, COD, pH, SS, T-N, T-P, Chl-a and Colon-bacterium observed from 1989 to 2013 in the Nakdong River. As a result of trends, SOM and RDA, the Mulgeum station was found that the water quality is improved, but caution is required in order to ensure safe water supply because concentrations in water quality were higher in the early spring(1~3 month) the most.

인공신경망 이론을 이용하여 강한 비선형성의 경향을 보이고 있는 강우-유출간의 관계를 모형화하기 위한 연구들은 예측뿐만이 아니라 대상자료들의 양상을 분류하여 그 특성을 분석하는 데에도 이용되고 있다. 이와 같은 패턴분류를 위한 SOM(Self-Organizing Map: SOM)의 연구 결과를 검토해보면 SOM 훈련을 위한 지도크기 및 배열의 결정은 SOM 성능에 큰 영향을 미치는 것으로 보고되고 있으나 지도크기 결정시 지도의 종방향 크기와 횡방향 크기를

The purpose of this study is to illustrate the characteristics of soil organic matter (SOM) and partition coefficient (KDOC). Humic substances (HS) from eight soils of varying properties were extracted by two different methods. The dissolved organic carbon (DOC) concentration was stabilized in 22hrs. The ratio of UV absorbance at 465nm and 665nm (E4/E6 ratio) for HS were similar pattern for 8 soils. The extraction with increasing pH increased dissolution of SOM. The ratio of organic carbon (OC) associated with HA and FA (the HA:FA ratio) was varied widely in accordance with the soils and was highly correlated to OC content(%) of the soils. In modeling metal speciation in soils and soil solutions, assumptions that all DOC in soil solution is associated with FA and that HA:FA ratio in SOM is constant have been made. The results of this study indicate that the validity of these assumptions is questionable. By sequential pH extraction, the KDOC showed in a linear correlation with pH.