• Journal of the Korea Computer Industry Society
• /
• v.2 no.2
• /
• pp.195-202
• /
• 2001

In this paper, we show a study on how to model a phoneme of which acoustic feature is changed according to both left-hand and right-hand phonemes. For this purpose, we make a comparative study on two kinds of algorithms; a unit reduction algorithm and decision tree modeling. The unit reduction algorithm uses only statistical information while the decision tree modeling uses statistical information and Korean acoustical information simultaneously. Especially, we focus on how to model context-dependent phonemes based on decision tree modeling. Finally, we show the recognition rate when context-dependent phonemes are obtained by the decision tree modeling.

• Communications for Statistical Applications and Methods
• /
• v.10 no.2
• /
• pp.471-478
• /
• 2003

In this paper, we reconsider the Kolmogorov-Smirnoff distance as a split criterion for binary decision trees and suggest an algorithm to obtain the Kolmogorov-Smirnoff distance more efficiently when the input variable have more than three categories. The Kolmogorov-Smirnoff distance is shown to have the property of exclusive preference. Empirical results, comparing the Kolmogorov-Smirnoff distance to the Gini index, show that the Kolmogorov-Smirnoff distance grows more accurate trees in terms of misclassification rate.

• Proceedings of the Korean Statistical Society Conference
• /
• 2004.11a
• /
• pp.105-110
• /
• 2004

Decision Analysis(DA) using Value Focused Thinking(VFT) can be an excellent process to deal with hard decisions. The intent of this research is to provide better understanding of the United States Air Force(USAF) officer retention problem. This thesis effort involves building a VFT model to find out more effective alternatives in retaining pilots and non pilots. This model, in conjunction with the results of the post analysis, shows an example of the application of a VFT approach to the USAF officer retention problem.

• Communications for Statistical Applications and Methods
• /
• v.10 no.3
• /
• pp.845-857
• /
• 2003

We explore the use of genetic programming to evolve decision trees directly for classification problems with both discrete and continuous predictors. We demonstrate that the derived hypotheses of standard algorithms can substantially deviated from the optimum. This deviation is partly due to their top-down style procedures. The performance of the system is measured on a set of real and simulated data sets and compared with the performance of well-known algorithms like CHAID, CART, C5.0, and QUEST. Proposed algorithm seems to be effective in handling problems caused by top-down style procedures of existing algorithms.

• Journal of the Korean Operations Research and Management Science Society
• /
• v.18 no.2
• /
• pp.45-55
• /
• 1993

In multiattribute decision making a decision maker (DM) can choose the best alternative if his/her multiattribute utility function and the joint probability distribution of outcomes are exactly known. This paper develops multiattribute stochastic-statistical dominance rules which can be applied to the situation when neither of them is known exactly, that is, when the DM cannot calculate the expected utility for each alternative. First, the notion of relative risk aversion is used dominance rules are developed to screen out dominated alternatives so that hi/she choose the best one among the remaining nondominated alternatives.

• Communications for Statistical Applications and Methods
• /
• v.7 no.3
• /
• pp.937-943
• /
• 2000

Data mining usually deal with undesigned massive data containing many variables for which their characteristics and association rules are unknown, therefore it is actually not easy to interpret the results of analysis. In this paper, it is shown that decision tree can be very useful in interpreting data mining prediction model using two real examples.

• Communications for Statistical Applications and Methods
• /
• v.9 no.2
• /
• pp.411-423
• /
• 2002

In this paper, we propose a split criterion for binary decision trees. The proposed criterion selects the optimal split by measuring the prediction success of the candidate splits at a given node. The criterion is shown to have the property of exclusive preference. Examples are given to demonstrate the properties of the criterion.

• The Korean Journal of Applied Statistics
• /
• v.19 no.2
• /
• pp.291-303
• /
• 2006

Data mining is the work to extract information from existing data file. So, the one of best important thing in data mining process is the quality of data to be used. In this thesis, we propose the data fusion technique using decision rule for data enrichment that one phase to improve data quality in KDD process. Simulations were performed to compare the proposed data fusion technique with the existing techniques. As a result, our data fusion technique using decision rule is characterized with low MSE or misclassification rate in fusion variables.

• Journal of Sasang Constitutional Medicine
• /
• v.13 no.2
• /
• pp.144-155
• /
• 2001

By the increasing concern about Sasang Constitution Medicine, its practical use is considered very important in disease prevention and medical treatment. However, the method of constitution classification is depending on the doctor's clinical trials because of the lack of the objective test criteria. This study is trying to improve the objectiveness of diagnosis using a new statistical method, decision tree. Decision tree method-a classification technique in the statistical analysis- was used to analyze the result of QSCCII instead of using discriminant analysis. As a result, 16 among 121 QSCCII questions was selected as important questions and 21 terminal nodes was built to classify the constitution. Using only 16 questions shown in the result of decision tree, we can diagnose and interpret the constitution easily and effectively.

• ETRI Journal
• /
• v.34 no.2
• /
• pp.184-189
• /
• 2012

In this paper, we propose a novel approach to statistical model-based voice activity detection (VAD) that incorporates a second-order conditional maximum a posteriori (CMAP) criterion. As a technical improvement for the first-order CMAP criterion in [1], we consider both the current observation and the voice activity decision in the previous two frames to take full consideration of the interframe correlation of voice activity. This is clearly different from the previous approach [1] in that we employ the voice activity decisions in the second-order (previous two frames) CMAP, which has quadruple thresholds with an additional degree of freedom, rather than the first-order (previous single frame). Also, a soft-decision scheme is incorporated, resulting in time-varying thresholds for further performance improvement. Experimental results show that the proposed algorithm outperforms the conventional CMAP-based VAD technique under various experimental conditions.