• The Korean Journal of Applied Statistics
• /
• v.28 no.3
• /
• pp.385-392
• /
• 2015

Principal component analysis (PCA), a dimension-reduction technique, is usually implemented after the variables are standardized when the measurement unit of variables are different. To standardize a variable we divide it by its standard deviation. But there is another way to transform a variable to be independent of its measurement unit. It is to divide it by its mean rather than standard deviation. Implementing PCA on standardized variables is equivalent to implementing PCA with a correlation matrix of original variables. Similarly, implementing PCA on the transformed variables divided by their means is equivalent to implementing PCA with a matrix related to the coefficients of variation of the original variables. We explain why we need to implement PCA on the variables transformed by their means.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.33 no.3
• /
• pp.22-29
• /
• 2010

Six sigma has been the most influential management innovation tool in order to achieve the customer's satisfaction and keep the competition in the age of limitless competition. The success in six sigma is to find the correct CTQ (Critical to Quality). QFD (Quality function deployment) is the efficient too ever created to tie product and service design decisions directly to customer wants and needs. One of the mistakes in QFD is to analyze using an inconsistent HOQ (House of quality) chart. An inconsistent HOQ chart is one in which the information from the correlation matrix is inconsistent with that from the relationship matrix. This study presents the consistency check and inconsistency check in case of failing the consistency check. Also we propose the procedures using the Limiting Probability in correlation matrix and the Median Rank in relationship matrix in order to be consistent in HOQ chart.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.12 no.7
• /
• pp.3128-3149
• /
• 2018

A number of effective methods for multiple-attribute group decision making (MAGDM) with interval-valued intuitionistic fuzzy numbers (IVIFNs) have been proposed in recent years. However, the different methods frequently yield different, even sometimes contradictory, results for the same problem. In this paper a novel criterion to determine the advantages and disadvantages of different methods is proposed. First, the decision-making process is divided into three parts: translation of experts' preferences, aggregation of experts' opinions, and comparison of the alternatives. Experts' preferences aggregation is considered the core step, and the quality of the collective matrix is considered the most important evaluation index for the aggregation methods. Then, methods to calculate the similarity measure, correlation, correlation coefficient, and energy of the intuitionistic fuzzy matrices are proposed, which are employed to evaluate the collective matrix. Thus, the optimal method can be selected by comparing the collective matrices when all the methods yield different results. Finally, a novel approach for aggregating experts' preferences with IVIFN is presented. In this approach, experts' preferences are mapped as points into two-dimensional planes, with the plant growth simulation algorithm (PGSA) being employed to calculate the optimal rally points, which are inversely mapped to IVIFNs to establish the collective matrix. In the study, four different methods are used to address one example problem to illustrate the feasibility and effectiveness of the proposed approach.

• Phonetics and Speech Sciences
• /
• v.1 no.1
• /
• pp.69-73
• /
• 2009

This paper proposes an efficient global covariance-based principal component analysis (GCPCA) for speaker identification. Principal component analysis (PCA) is a feature extraction method which reduces the dimension of the feature vectors and the correlation among the feature vectors by projecting the original feature space into a small subspace through a transformation. However, it requires a larger amount of training data when performing PCA to find the eigenvalue and eigenvector matrix using the full covariance matrix by each speaker. The proposed method first calculates the global covariance matrix using training data of all speakers. It then finds the eigenvalue matrix and the corresponding eigenvector matrix from the global covariance matrix. Compared to conventional PCA and Gaussian mixture model (GMM) methods, the proposed method shows better performance while requiring less storage space and complexity in speaker identification.

• Journal of the Institute of Convergence Signal Processing
• /
• v.9 no.4
• /
• pp.304-312
• /
• 2008

Orthogonal codes like Walsh and Golay codes may have large correlation value when they are not synchronized, hence they are seldom used in asynchronous CDMA systems. Wysocki[1] showed that by multiplying the original Walsh-Hadamard matrix with an orthogonal transformation matrix the resultant matrix sustains orthogonality between row vectors and their cross-correlation can be reduced. Soberly and Wysocki[2] proposed similar scheme on Golay codes. This implies that using the proper orthogonal transformation cross-correlation of Walsh and Golay codes can be reduced, and the transformed codes can be used for user separation in the CDAM reverse link. In this paper we discuss cross-correlation related parameters which affect the performance of an asynchronous CDMA link, and we investigate the correlation properties of the transformed codes. When we designed orthogonal transformation matrices for Walsh and Golay codes, we minimized the maximum value of aperiodic cross-correlation of the codes ($ACC_{max}$) or the mean square value of the aperiodic cross-correlation($R_{cc}$) with preserving the orthogonality of the modified codes. We also evaluate the asynchronous CDMA system that uses the transformed Walsh and Golay codes.

• The Journal of Fisheries Business Administration
• /
• v.15 no.1
• /
• pp.58-80
• /
• 1984

The classification of islands is prerequisite for establishing a development policy to vitalize many-sided function of islands. We try to classify the 440 inhabited islands which exist in Jeon-Nam area and Kyong-Nam area by means of PCA. PCA begins with making correlation matrix of orignal variables. From this matrix we can comprehend the rough relationships between two variables. Next, we look for the eigenvalues which are roots of characteristic equation of correlation matrix. The number of eigenvalues is equal to that of original variables. We choose the largest eigenvalue λ$_1$among them and then look for the eigenvector of correlation matrix corresponding to the largest eigenvalue. Linear combination of eigenvector obtained above and original variables is namely first Principal Component (PC). Using an eigenvalue criterion(λ$\geq$ 1), we choose 3 PCs in Jeon-Nam area and 2 PCs in Kyong-Nam area. But we decide to consider only two PCs in both areas to faciliate a comparative analysis. Now, loss of information is 31.7% in Jeon-Nam area and 26.64% in Kyong-Nam area. PCs extracted by preceding procedure have characteristics as follows. The first PC relates to aggregate size of islands in case of both areas. The second PC relates to income per household, factors of agricultural production and factors of fisheries production in Jeon-Nam area, but in Kyong-Nam area it means distance from island and income per household. A classification of islands can be attained by plotting component scores of each island in graph used two PCs as axes and grouping similiar islands. 6 groups are formed in Jeon-Nam area and 5 groups in Kyong-Nam area. The result of this study in kyong-Nam area accords with prior result of study.

• Tunnel and Underground Space
• /
• v.18 no.6
• /
• pp.408-417
• /
• 2008

Rock mass classification methods such as RMR and Q system have different characteristics each other in parameters considered and applications, and so it is very important to prescribe the relationship between parameters for the analysis of correlativity of these methods. With the Held data of RMR and Q estimation in road construction sites, the acquaintance relations between RMR and Q of rock mass classifications are analyzed. The correlation equations between parameters of RMR and Q, matrix of correlation coefficients and the generalized form of acquaintance relation matrix are derived. This acquaintance relation matrix can be further extended to the form of generalized acquaintance relation network, and could be used to analyze the correlativity and to enhance the utility of common rock mass classification methods.

• Asian Pacific Journal of Cancer Prevention
• /
• v.16 no.15
• /
• pp.6749-6751
• /
• 2015

Background: Laryngeal cancer is an important malignancy in head and neck area and squamous cell carcinoma (SCC) is the most common type accounting for 95% of cases. Increase in matrix metalloproteinases (MMPs) in different tumors and their correlation with tumor invasiveness has been documented. However, most studies have evaluated MMP-2 and MMP-9 expression and few have evaluated serum levels. The aim of current study was to evaluate serum levels in patients with laryngeal SCC compared to normal subjects and assess any relation with tumor clinicopathological findings. Materials and Methods: In this case control study, 20 patients with oral SCC and 20 healthy subjects were included. Serum levels of MMP-2 and MMP-9 were compared between groups and correlations with findings including grade (T) and node involvement (N) were evaluated. Results: Patients with laryngeal SCC had significantly higher serum levels of MMP-2 (p=0.01) and MMP-9 (p=0.03) compared to healthy subjects. Patients with higher T stage (T3,4) had significantly higher MMP-2 (p=0.04) and MMP-9 (p=0.01). There was significant positive correlation between serum levels of MMP-2 with T stage (r=0.45, p=0.04) and lymph node involvement (r=0.563, p=0.01) and between levels of MMP-9 with T stage (r=0.527, p=0.01). Conclusions: Our results showed that compared to healthy subjects, both MMP-2 and MMP-9 are significantly increased in serum of laryngeal SCC cases. MMP-2 was correlated with lymph node involvement while MMP-9 has stronger correlation with T stage compared to MMP-2.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.21 no.1
• /
• pp.79-86
• /
• 2021

In order to increase the probability of success in big data projects, quantified techniques are required to analyze the root cause of risks from complex causes and establish optimal countermeasures. To this end, this study measures risk factors and relationships through SNA analysis and presents a way to respond to risks based on them. In other words, it derives a dependency network matrix by utilizing the results of correlation analysis between risk groups in the big data projects presented in the preliminary study and performs SNA analysis. In order to derive the dependency network matrix, partial correlation is obtained from the correlation between the risk nodes, and activity dependencies are derived by node by calculating the correlation influence and correlation dependency, thereby producing the causal relationship between the risk nodes and the degree of influence between all nodes in correlation. Recognizing the root cause of risks from networks between risk factors derived through SNA between risk factors enables more optimized and efficient risk management. This study is the first to apply SNA analysis techniques in relation to risk management response, and the results of this study are significant in that it not only optimizes the sequence of risk management for major risks in relation to risk management in IT projects but also presents a new risk analysis technique for risk control.

• International conference on construction engineering and project management
• /
• 2015.10a
• /
• pp.210-212
• /
• 2015

The growth-share matrix is a portfolio planning tool developed by the Boston Consulting Group (BCG) to assist competitive positioning in the international market including those in the construction industry. This matrix is helpful in balancing the firm's cash-flow, and it can suggest strategic directions for each business unit. However, its effectiveness and applicability have long been debated in the academic field due to the complex and unique industrial context of construction. To solve the dispute, this research clarifies the applicability of theories underlying the growth-share matrix to the construction industry. Empirical research based on actual financial data of Korean construction firms is adopted for the statistical analysis including one-way analysis of variance and correlation analysis. The results of this research show that empirical findings on the relationship between performance variables. In this context, this research can provide important insights on the concept of portfolio management in the construction industry.