• Journal of the Korean Institute of Intelligent Systems
• /
• v.15 no.1
• /
• pp.81-86
• /
• 2005

In this study, we introduce a new category of fuzzy inference systems based on information granulation to carry out the model identification of complex and nonlinear systems. Informally speaking, information granules are viewed as linked collections of objects (data, in particular) drawn together by the criteria of proximity, similarity, or functionality Granulation of information with the aid of Hard C-Means (HCM) clustering help determine the initial parameters of fuzzy model such as the initial apexes of the membership functions and the initial values of polynomial functions being used in the premise and consequence part of the fuzzy rules. And the initial parameters are tuned effectively with the aid of the genetic algorithms(GAs) and the least square method (LSM). An aggregate objective function with a weighting factor is also used in order to achieve a balance between performance of the fuzzy model. The proposed model is evaluated with using a numerical example and is contrasted with the performance of conventional fuzzy models in the literature.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.43 no.4 s.310
• /
• pp.50-57
• /
• 2006

In Electronic Commerce, the latest most of the personalized recommender systems have applied to the collaborative filtering technique. This method calculates the weight of similarity among users who have a similar preference degree in order to predict and recommend the item which hits to propensity of users. In this case, we commonly use Pearson Correlation Coefficient. However, this method is feasible to calculate a correlation if only there are the items that two users evaluated a preference degree in common. Accordingly, the accuracy of prediction falls. The weight of similarity can affect not only the case which predicts the item which hits to propensity of users, but also the performance of the personalized recommender system. In this study, we verify the improvement of the prediction accuracy through an experiment after observing the rule of the weight of similarity applying Vector similarity, Entropy, Inverse user frequency, and Default voting of Information Retrieval field. The result shows that the method combining the weight of similarity using the Entropy with Default voting got the most efficient performance.

• Spatial Information Research
• /
• v.21 no.5
• /
• pp.15-22
• /
• 2013

The purpose of this work is to investigate correlation between $CO_2$ concentration and NDVI (Normalized Difference Vegetation Index) in North East Asia. Geographically weighted regression techniques were used to evaluate the spatial relationships between GOSAT (Greenhouse Observing SATellite) $CO_2$ measurement and MODIS (Moderate Resolution Imaging Spectroradiometer) vegetation index. The results reveals that $CO_2$ concentration to be negatively associated with NDVI. The analysis of Global Morans' I index and Anselin Local Morasn's I showed spatial autocorrelation between the overall spatial pattern of $CO_2$ and NDVI. Ultimately, there were clustered patterns in both data sets. The results show that carbon dioxide concentration shows non-random distribution patterns in relation to NDVI clusters, which proves that intense development activities such as deforestation are influencing carbon dioxide emission across the area of analysis. However, as the concentration of carbon dioxide varies depending on a variety of factors such as artificial sources, plant respiration, and the absorption and discharge of the ocean, follow-up studies are required to evaluate the correlations among more related variables.

• Journal of the Economic Geographical Society of Korea
• /
• v.11 no.4
• /
• pp.564-579
• /
• 2008

It is often challenging to evaluate local spatial association among onedimensional vectors generally representing paired-location events where two points are physically or functionally connected. This is largely because of complex process of such geographic phenomena itself and partially representational complexity. This paper addresses an alternative way to identify spatially autocorrelated paired-location events (or vectors) at a local scale. In doing so, we propose a statistical algorithm combining univariate point pattern analysis for evaluating local clustering of origin-points and similarity measure of corresponding vectors. For practical use of the suggested method, we present an empirical application using transactions data in a local housing market, particularly recorded from 2004 to 2006 in Franklin County, Ohio in the United States. As a result, several locally characterized similar transactions are identified among a set of vectors showing various local moves associated with communities defined.

• Journal of the Korean Society of Costume
• /
• v.24
• /
• pp.217-233
• /
• 1995

The ultimate purpose of this study is to suggest the most effective marketing strategy for the clothing consumer market based on the new generation consumer's clothing selection behavior analysis. In this thesis, it is appempted to make a progress in the new gen-eration consumer's clothing life style types, in clothing purchase behavior analysis among the clothing life style, and also in the marketing strategy for marketers. The subjects selected for the final analysis are 412 the new gerneration women of age 20 thru 34 in seoul and satellite town area. Data were processed the spss package program. As for the analytic method, factor analysis, clustering analysis, XCross-tubulation, F-test with ANOVA, frequency and percentage were applied in the survey. The major findings are as following : life style is classified into four types : The characteristic fashion-directory type(25.7%) ; The reason traditional type(9.0%) ; The sen-sitivity fashion-following type(11.0%) ; The community brand-conscious type(54.3%). 2 Clothing life style types characteristic of the new generation consumer proved that clothing life style types are a significant difference according to the life style, the fashion consciousness and the average monthly spend-ing on clothing. 3. There is an important discrimination according to the clothing life style types in their clothing purchase behavior such as infor-mation usage, clothing choice criterion and brand loyalty. 4. Based on the result of our analysis and the review of literature, the marketing strategy is suggested that characteristic and new design development is efficient way to consumer's purchase need. Therefore apparel industary which pursue an added value must frame marketing strategy on the basis of the target consumer's sensitivity characteristic according to the life style and fashion consciousness.

• Molecular & Cellular Toxicology
• /
• v.5 no.1
• /
• pp.58-66
• /
• 2009

The 2-year rodent carcinogenicity test involves long-term, repetitive dosing of animals that is both time consuming and expensive. Alternative approaches have been attempted using specific transgenic or knockout mice or toxicogenomics to predict carcinogenicity without conducting a 2-year rodent test. In addition, toxicogenomic analysis of carcinogen-treated animals could also enhance our understanding of molecular mechanisms and aid in the diagnosis of acute toxicity induced by carcinogens. Therefore, we investigated transcription profiles after administering the carcinogens 4,4-dimethylformamide (DMF) and 4-biphenylamine (ABP). BALB/c male mice were treated once with DMF (650 mg/kg i.p.) or ABP (120 mg/kg p.o.). Standard blood biochemistry and histological changes were observed. Gene expression profiles in the livers of mice treated with either vehicle or the carcinogens were analyzed using the Affymetrix $GeneChip^{(R)}$ assay. In all, 1,474 differentially expressed genes in DMF- or ABP-treated mice were identified as being either up- or down-regulated over 1.5-fold (P< 0.01), and these genes were analyzed using hierarchical clustering and Ingenuity Pathways Analysis. Of these, 107 genes were consistently regulated in both carcinogen-treated groups. Genes associated with cancer were upregulated (Por, S100a10, Tes, Ctcf, Ddx21, Eapp, Nel, and Pa2g4) or downregulated (Cbs and Gch1). Toxicological function analysis also identified genes involved in organ toxicity, including hepatotoxicity. These data may help to identify molecular markers for acute hepatotoxicity induced by carcinogens.

• Journal of Preventive Medicine and Public Health
• /
• v.41 no.1
• /
• pp.1-9
• /
• 2008

Objectives : This study evaluated knowledge structure and its effect factor by analysis of co-author and keyword networks in Korea's preventive medicine sector. Methods : The data was extracted from 873 papers listed in the Journal of Preventive Medicine and Public Health, and was transformed into a co-author and keyword matrix where the existence of a 'link' was judged by impact factors calculated by the weight value of the role and rate of author participation. Research achievement was dependent upon the author's status and networking index, as analyzed by neighborhood degree, multidimensional scaling, correspondence analysis, and multiple regression. Results : Co-author networks developed as randomness network in the center of a few high-productivity researchers. In particular, closeness centrality was more developed than degree centrality. Also, power law distribution was discovered in impact factor and research productivity by college affiliation. In multiple regression, the effect of the author's role was significant in both the impact factor calculated by the participatory rate and the number of listed articles. However, the number of listed articles varied by sex. Conclusions : This study shows that the small world phenomenon exists in co-author and keyword networks in a journal, as in citation networks. However, the differentiation of knowledge structure in the field of preventive medicine was relatively restricted by specialization.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.20 no.2
• /
• pp.81-86
• /
• 2020

Ambient backscatter communication has a drawback in which the transmission power is limited because the data is transmitted using the ambient RF signal. In order to improve transmission efficiency between transceiver, a channel estimator capable of estimating channel state at a receiver is needed. In this paper, we consider the K-means algorithm to improve the performance of the channel estimator based on EM algorithm. The simulation uses MSE as a performance parameter to verify the performance of the proposed channel estimator. The initial value setting through K-means shows improved performance compared to the channel estimation method using the general EM algorithm.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.48 no.2
• /
• pp.125-133
• /
• 2011

For any pattern matching based algorithm in WLAN environment, the characteristics of signal to noise ratio(SNR) to multiple access points(APs) are utilized to establish database in the training phase, and in the estimation phase, the actual two dimensional coordinates of mobile unit(MU) are estimated based on the comparison between the new recorded SNR and fingerprints stored in database. As fingerprinting method, k-nearest neighbor(KNN) has been widely applied for indoor location in wireless location area networks(WLAN), but its performance is sensitive to number of neighbors k and positions of reference points(RPs). So intuitive fuzzy c-means(IFCM) clustering algorithm is applied to improve KNN, which is the KNN/IFCM hybrid algorithm presented in this paper. In the proposed algorithm, through KNN, k RPs are firstly chosen as the data samples of IFCM based on signal to noise ratio(SNR). Then, the k RPs are classified into different clusters through IFCM based on SNR. Experimental results indicate that the proposed KNN/IFCM hybrid algorithm generally outperforms KNN, KNN/FCM, KNN/PFCM algorithm when the locations error is less than 2m.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.48 no.2
• /
• pp.109-115
• /
• 2011

As fingerprinting method, k-nearest neighbor(KNN) has been widely applied for indoor location in wireless location area networks(WLAN), but its performance is sensitive to number of neighbors k and positions of reference points(RPs). So artificial neural network(ANN) clustering algorithm is applied to improve KNN, which is the KNN/ANN hybrid algorithm presented in this paper. For any pattern matching based algorithm in WLAN environment, the characteristics of signal to noise ratio(SNR) to multiple access points(APs) are utilized to establish database in the training phase, and in the estimation phase, the actual two dimensional coordinates of mobile unit(MU) are estimated based on the comparison between the new recorded SNR and fingerprints stored in database. In the proposed algorithm, through KNN, k RPs are firstly chosen as the data samples of ANN based on SNR. Then, the k RPs are classified into different clusters through ANN based on SNR. Experimental results indicate that the proposed KNN/ANN hybrid algorithm generally outperforms KNN algorithm when the locations error is less than 2m.