• Journal of Korean Society for Quality Management
• /
• v.47 no.2
• /
• pp.271-281
• /
• 2019

Purpose: The purpose of this study is to suggest a more efficient distance measure taking into account the relationship between categorical variables for categorical data cluster analysis. Methods: In this study, the association-based dissimilarity was employed to calculate the distance between two categorical data observations and the distance obtained from the association-based dissimilarity was applied to the PAM cluster algorithms to verify its effectiveness. The strength of association between two different categorical variables can be calculated using a mixture of dissimilarities between the conditional probability distributions of other categorical variables, given these two categorical values. In particular, this method is suitable for datasets whose categorical variables are highly correlated. Results: The simulation results using several real life data showed that the proposed distance which considered relationships among the categorical variables generally yielded better clustering performance than the Hamming distance. In addition, as the number of correlated variables was increasing, the difference in the performance of the two clustering methods based on different distance measures became statistically more significant. Conclusion: This study revealed that the adoption of the relationship between categorical variables using our proposed method positively affected the results of cluster analysis.

• Journal of Korean Society for Quality Management
• /
• v.47 no.3
• /
• pp.537-552
• /
• 2019

Purpose: The development of information technology makes it easy to utilize high-dimensional categorical data. In this regard, the purpose of this study is to propose a novel method to select the proper categorical variables in high-dimensional categorical data. Methods: The proposed feature selection method consists of three steps: (1) The first step defines the goodness-to-pick measure. In this paper, a categorical variable is relevant if it has relationships among other variables. According to the above definition of relevant variables, the goodness-to-pick measure calculates the normalized conditional entropy with other variables. (2) The second step finds the relevant feature subset from the original variables set. This step decides whether a variable is relevant or not. (3) The third step eliminates redundancy variables from the relevant feature subset. Results: Our experimental results showed that the proposed feature selection method generally yielded better classification performance than without feature selection in high-dimensional categorical data, especially as the number of irrelevant categorical variables increase. Besides, as the number of irrelevant categorical variables that have imbalanced categorical values is increasing, the difference in accuracy between the proposed method and the existing methods being compared increases. Conclusion: According to experimental results, we confirmed that the proposed method makes it possible to consistently produce high classification accuracy rates in high-dimensional categorical data. Therefore, the proposed method is promising to be used effectively in high-dimensional situation.

• Communications for Statistical Applications and Methods
• /
• v.30 no.6
• /
• pp.577-587
• /
• 2023

Conventional categorical data imputation techniques, such as mode imputation, often encounter issues related to overestimation. If the variable has too many categories, multinomial logistic regression imputation method may be impossible due to computational limitations. To rectify these limitations, we propose a two-stage imputation method. During the first stage, we utilize the Boruta variable selection method on the complete dataset to identify significant variables for the target categorical variable. Then, in the second stage, we use the important variables for the target categorical variable for logistic regression to impute missing data in binary variables, polytomous regression to impute missing data in categorical variables, and predictive mean matching to impute missing data in quantitative variables. Through analysis of both asymmetric and non-normal simulated and real data, we demonstrate that the two-stage imputation method outperforms imputation methods lacking variable selection, as evidenced by accuracy measures. During the analysis of real survey data, we also demonstrate that our suggested two-stage imputation method surpasses the current imputation approach in terms of accuracy.

• Journal of the Korean Data and Information Science Society
• /
• v.21 no.6
• /
• pp.1353-1359
• /
• 2010

Basic objective in cluster analysis is to discover natural groupings of items. In general, clustering is conducted based on some similarity (or dissimilarity) matrix or the original input data. Various measures of similarities between objects are developed. In this paper, we consider a clustering of huge categorical real data set which shows the aspects of time-location-activity of Korean people. Some useful similarity measure for the data set, are developed and adopted for the categorical variables. Hierarchical and nonhierarchical clustering method are applied for the considered data set which is huge and consists of many categorical variables.

• IE interfaces
• /
• v.12 no.4
• /
• pp.551-556
• /
• 1999

In this paper, we compare the performance of three data mining classification algorithms(neural network, decision tree, logistic regression) in consideration of various characteristics of categorical input and output data. $2^{4-1}$. 3 fractional factorial design is used to simulate the comparison situation where factors used are (1) the categorical ratio of input variables, (2) the complexity of functional relationship between the output and input variables, (3) the size of randomness in the relationship, (4) the categorical ratio of an output variable, and (5) the classification algorithm. Experimental study results indicate the following: decision tree performs better than the others when the relationship between output and input variables is simple while logistic regression is better when the other way is around; and neural network appears a better choice than the others when the randomness in the relationship is relatively large. We also use Taguchi design to improve the practicality of our study results by letting the relationship between the output and input variables as a noise factor. As a result, the classification accuracy of neural network and decision tree turns out to be higher than that of logistic regression, when the categorical proportion of the output variable is even.

• Communications for Statistical Applications and Methods
• /
• v.13 no.2
• /
• pp.297-307
• /
• 2006

Most methods for describing the relationship among random variables require specific probability distributions and some assumptions of random variables. The mutual information based on the entropy to measure the dependency among random variables does not need any specific assumptions. And the redundancy which is a analogous version of the mutual information was also proposed. In this paper, the redundancy and mutual information are explored to multi-dimensional categorical data. It is found that the redundancy for categorical data could be expressed as the function of the generalized likelihood ratio statistic under several kinds of independent log-linear models, so that the redundancy could also be used to analyze contingency tables. Whereas the generalized likelihood ratio statistic to test the goodness-of-fit of the log-linear models is sensitive to the sample size, the redundancy for categorical data does not depend on sample size but its cell probabilities itself.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.37 no.4
• /
• pp.168-176
• /
• 2014

The enterprise life cycle derived from the product life cycle consists of introduction, growth, maturity and decline. The enterprise tries to reach the growth stage early and stay at the maturity stage stably through expanding its businesses and investing for the new technology. The public enterprise is not different but its life cycle is more prone to be affected by the national development and policy. A typical example can be found in the case of the quasi market SOC public enterprise which spends massive amount of fund to provide social infrastructure. After the fulfillment of its mandated mission it is exposed to the pressure of a merger or a closure usually because large portion of the debt is directly linked to the national financial stability and credit ratings. This research is focused on the variables that influence the life cycle of the quasi market SOC public Enterprise for its future competitiveness is in connection with its normalization, advancement and rationalization. In this respect, categorical variables system centering on public characteristics and profitability drew eight categorical variables such as policy outcomes, public benefit, finance and business values etc.

• Journal of the Korean Statistical Society
• /
• v.5 no.1
• /
• pp.29-34
• /
• 1976

The techniques about the analysis of variance for quantitative variables have been well-developed. But when the variable is categorical, we must switch to a completely different set of varied techniques. R.J. Light and B.H. Margolin presented one kind of techniques for categorical data in their paper, where there are G unordered experimental groups and I unordered response categories.

• Journal of the Korean Data and Information Science Society
• /
• v.7 no.2
• /
• pp.219-226
• /
• 1996

Basic objective in cluster analysis is to discover natural groupings of items or variables. In general, variable clustering was conducted based on some similarity measures between variables which have binary characteristics. We propose a variable clustering method when variables have more categories ordered in some sense. We also consider some measures of association as a similarity between variables. Numerical example is included.

• Journal of Korean Society for Quality Management
• /
• v.32 no.3
• /
• pp.109-125
• /
• 2004

This paper deals with an optimal process control problem in production of hull structural steel plate with high defective rate. The main quality characteristic(dependent variable) is the internal quality(defect) of plates and is dependent on process parameters(independent variables). The dependent variable(quality characteristics) has three categorical ordinal data and there are 35 independent variables(29 continuous variables and 6 categorical variables). In this paper, we determine the main factors and to develop the mathematical model between internal quality predicted probabilities and the main factors. Secondly, we find out the optimal process condition of main factors through analysis of variance(ANOVA) using simulation. We consider three models to obtain the main factors and the optimal process condition: linear, quadratic, error models.