의사결정나무분석과 로지스틱 회귀분석을 이용한 우울 예측요인 비교연구

윤지선 ( Youn Ji Sun ) 한국사회복지경영학회 2020 사회복지경영연구 Vol.7 No.2

본 연구는 노년기 삶의 질을 저해하는 우울증에 대한 관심으로부터 수행되었다. 의사결정나무(decision tree) 분석을 활용하여 노인의 우울 요인을 분류 및 예측하고, 이를 로지스틱 회귀분석 결과와 비교하여 예측 정확성을 정의하는 서술적 조사연구이다. 연구대상자는 국민연금연구원의 국민노후보장패널(KReIS) data 중, 7차 개인조사에 참가한 65세 이상 노인 총 2,096명이다. 자료분석은 SPSS 23.0 프로그램을 이용하여 기술통계, 교차분석, Roc Curve, 의사결정나무 분석, 로지스틱 회귀분석을 하였다. 연구결과, 의사결정나무 분석에서 우울 예측요인은 일상 및 사회생활 제한과 주관적 경제 불만족으로 나타났다. 로지스틱 회귀분석에서는 일상 및 사회생활 제한과 주관적 경제 불만족, 대인관계 불만족으로 나타났다. 노인의 우울에 영향을 미치는 예측력을 로지스틱 회귀분석과 의사결정나무 분석을 통해 비교한 결과, 우울을 예측하는 민감도는 로지스틱 회귀분석(44.4%)이 의사결정나무 분석(33.6%) 보다 높게 나타났다. 또한 실제 우울을 예측하는 특이도는 의사결정나무 분석(91.9%)이 로지스틱 회귀분석(86.3%) 보다 높은 것으로 나타났다. 분류정확도는 로지스틱 회귀분석(71.6%)이 의사결정나무 분석(71.4%)보다 조금 높게 나타났다. 연구결과를 기초로 두 기법의 예측 및 분류도 구로서의 유용성 판단은 민감도와 분류 정확도가 더 높게 나타난 로지스틱 회귀분석방법이 노인의 우울 예측모형을 구축하는데 더 유용한 자료로 사용될 수 있을 것으로 사료된다. 반면, 의사결정나무 분석은 분석의 정확도보다는 분석과정의 특정 경로설명이 필요한 경우에 유용하게 사용될 수 있을 것으로 보인다. This study was carried out from the interest in depression, which undermines the quality of life in old age, which has been extended by life expectancy. It is a descriptive investigation study that utilizes decision tree analysis with data mining technique to classify and predict depression factors of the elderly, and compare them with logistic regression results to define prediction accuracy. Among the data of the National Pension Research Institute's Korea National Age Security Panel(KReIS), a total of 2,096 senior citizens aged 65 or older participated in the seventh personal survey conducted in 2017. The data analysis was performed using the SPSS 23.0 program, including technical statistics, cross-analysis, logistic regression, Loc Curve, and decision tree analysis. The results of the study showed that the factors of depression prediction in decision tree analysis were daily and social life restriction and subjective economic dissatisfaction. Logistic regression showed limitations in daily and social life, subjective economic dissatisfaction and interpersonal dissatisfaction. Comparing the predictive power that affects the depression of the elderly through logistic regression and decision tree analysis, the sensitivity of predicting depression was higher than that of the decision tree(33.6%). In addition, the specificity of predicting actual depression was higher than that of logistic regression(86.3%) with decision tree analysis(91.9%). Classification accuracy was slightly higher than logistic regression(71.6%) in decision tree analysis(71.4%). Based on the results of the study, it is estimated that the logistic regression method, which shows higher sensitivity and accuracy of classification, can be used as more useful data to build a depression prediction model for the elderly. On the other hand, decision tree analysis may be useful when specific path descriptions of the analysis process are needed rather than the accuracy of the analysis.

Towards A Better Approach for Mastering Industrial Risks From Modeling Accidental Process to Integrating Safety Analysis Techniques Supporting the Identification of Intelligent Safety Decision

Dahmane Djamal,Bahmed Lylia,Bensiali Abdelkarim 보안공학연구지원센터 2015 International Journal of u- and e- Service, Scienc Vol.8 No.3

This paper introduces a research aiming at the development of a new approach to mastering industrial risks and prevent accident scenario. Starting from modeling and analyzing accidental process to understand the causes of accidents using quantitative risk assessment and reliability, they are essential issues in modern safety to make reliable decision, and it is new approach used for risk management to identify accident scenarios that may occur at their facilities which are sources of damage. Risk assessment of safety instrumented systems is approaches designed primarily to reduce the existing risk inherent in engineering system to a level considered tolerable and maintain it over time. In this study, the reliability of quantitative risk assessment using fuzzy sets based on event tree analysis and layer of protection analysis is the model proposed to deal with inaccuracy and uncertainty of data, The model proposed to determine the severity of the scenario and determine the safety integrity level SIL using Fuzzy Sets theory. The results which have got by this model is more motivate to deal with uncertainty of which considered as complementary for logical and arithmetic computation. As the accident is a chain of failure events, each related to its (causal) event or events, the early detection and diagnosis of faults in processes is very important, we use Fault Tree Analysis to show the possible malfunctions by enumerating the suspect components and their respective failure modes. Fault diagnosis when error occurs is performed by engineers and analysts performing extensive examination of all data feedback from the system and decisions are made based on threshold values by using fault trees. Since those decision-making tasks are safety critical and must be done promptly, the engineers who manually analyze the data are facing time challenge. To automate this process, we present in this approach to use decision trees to capture the contents of fault trees and detect faults by running the telemetry data through the decision trees in real time. Decision trees are the binary trees built from data samples and can classify the objects into different classes, the decision trees can classify different fault events or normal events. Given a set of data samples, decision trees can be built and trained, and then by running the new data through the trees, classification and prediction can be made.

의사결정나무 분석을 통한 한국 중고령자의 점진적 은퇴의사결정에 관한 연구

임은정(Eunjung Lim),정순희(Soonhee Jeong) 한국FP학회 2015 Financial Planning Review Vol.8 No.3

본 연구는 한국고용정보원에서 발표하는 『한국 고령화 패널 연구(Korea Longitudinal Study of Ageing)』의 2012년 자료를 이용하여 모수적 가정을 하는 기존의 통계분석의 한계점을 보완할 수 있는 데이터마이닝 기법 중 의사결정나무 모형을 구축함으로써 한국 중고령자의 점진적 은퇴 결정요인 및 규칙을 도출하여 은퇴의사결정을 설명하는데 연구의 목적을 두었다. 연구의 주요결과는 다음과 같다. 첫째, 최적의 의사결정나무 모형 구축 결과, 현 일자리에 대한 기술적합도를 매우 낮게 인식하고, 구직 시 제시되는 일자리의 적은 수입으로 인한 어려움을 덜 느끼며, 은퇴계획연령이 68.6세 이상 혹은 66.1세 미만인 경우 중고령자들은 점진적 은퇴를 선택하는 것으로 나타났다. 둘째, 완전히 성장한 의사결정나무 모형구축 결과, 최적의 의사결정나무에서 도출된 매우 낮은 기술적합도, 적은 수입으로 인한 구직시의 어려움, 68.6세 이상 혹은 66.1세 미만의 계획된 은퇴연령의 요인 이외에도, 최종학력이 중졸이상이며, 지난해의 가구총소득이 1,100만원 미만인 경우 점진적 은퇴를 선택하는 것으로 나타났다. 반면 기술적합도를 높게 느끼는 경우에는 현 일자리에 대한 높은 만족도, 68.6세 이상의 은퇴계획연령이 점진적 은퇴의 선택요인이었다. 이를 통해 그 어떤 요인보다도 중고령자의 직업환경과 구직환경, 그리고 계획한 은퇴연령이 주요한 영향요인임을 알 수 있었으며, 특히 은퇴계획연령의 경우 현재 평균 퇴직연령인 52.6세 및 법상에서 명시된 정년연령인 60세와도 차이가 나고 있음을 확인 할 수 있었다. 셋째, 구축된 의사결정나무 모형의 우수성을 검증하기 위해 다항 로짓 회귀분석의 결과와 비교해본 결과 의사결정나무의 전체 에러율이 2.30%, 다항 로짓 회귀분석의 전체 에러율이 5.40%로 나타나 의사결정나무 모형이 영향요인 도출 및 규칙 설명에 더 우수함을 확인할 수 있었다. 또한 네 가지 종속변수의 실제값과 분석결과로 나타난 분류값 간의 오차에 대한 차이를 살펴보기 위해 실시한 맥네머 검정(McNemar Test) 결과에서도 유의확률이 .000으로 의사결정나무와 다항 로짓 회귀분석의 분류 정확도에는 통계적으로 유의미한 차이가 나타나고 있음을 확인하였다. This study aimed at deriving determinants and rules of retirement process of the Korean middle-old aged and describing retirement decision-making by constructing a decision tree model from data mining techniques for complementing the limitation of the existing statistical analysis. The study used the data from the 2011 Korea Longitudinal Study of Ageing collected by Korea Employment Information Service. Main research findings are as follows. Firstly, as a result of constructing the optimal decision tree model, it showed that the middle to old aged selected gradual retirement in case that they recognized a very low technological suitability on the present job, did not have much difficulties from small income when they searched for a job and that they planned retirement age more than 68.6 or less than 66.1. Secondly, as a result of constructing a completely grown decision tree model, difficulties from small income during job search and planned retirement age more than 68.6 or less than 66.1 derived from the optimal decision tree. This revealed that above middle school graduates, individuals who had under 11 million won in the last year’s family gross income selected gradual retirement. On the contrary, individuals who felt a higher technological suitability, highly satisfied with the present job and who planned retirement aged more than 68.6 selected gradual retirement. It means that vocational and employment environments and planned retirement age are significant factors. Particularly, it was identified that the difference existed between planned retirement age and the current average retirement age or the retirement age by the law. Thirdly, in order to verify superiority of the constructed decision tree model, the results was compared with the findings of multinomial logistic regression It showed the total error rate was 2.30% for the decision tree and 5.40% for multinomial logistic regression. So, it was identified that decision tree model is superior for deriving significant factors and explaining rules. Besides, McNemar Test was conducted to examine difference of errors between the actual value of 4 dependant variables and the classification value from analysis. As a result, with the significance probability of .000, it showed that there is statistically significant difference between classification accuracies for the decision tree and the multinomial logistic regression.

의사결정나무 분석을 통한 한국 중고령자의 점진적 은퇴의사결정에 관한 연구

임은정,정순희 한국FP학회 2015 Financial Planning Review Vol.8 No.4

This study aimed at deriving determinants and rules of retirement process of the Korean middle-old aged and describing retirement decision-making by constructing a decision tree model from data mining techniques for complementing the limitation of the existing statistical analysis. The study used the data from the 2011 Korea Longitudinal Study of Ageing collected by Korea Employment Information Service. Main research findings are as follows. Firstly, as a result of constructing the optimal decision tree model, it showed that the middle to old aged selected gradual retirement in case that they recognized a very low technological suitability on the present job, did not have much difficulties from small income when they searched for a job and that they planned retirement age more than 68.6 or less than 66.1. Secondly, as a result of constructing a completely grown decision tree model, difficulties from small income during job search and planned retirement age more than 68.6 or less than 66.1 derived from the optimal decision tree. This revealed that above middle school graduates, individuals who had under 11 million won in the last year’s family gross income selected gradual retirement. On the contrary, individuals who felt a higher technological suitability, highly satisfied with the present job and who planned retirement aged more than 68.6 selected gradual retirement. It means that vocational and employment environments and planned retirement age are significant factors. Particularly, it was identified that the difference existed between planned retirement age and the current average retirement age or the retirement age by the law. Thirdly, in order to verify superiority of the constructed decision tree model, the results was compared with the findings of multinomial logistic regression It showed the total error rate was 2.30% for the decision tree and 5.40% for multinomial logistic regression. So, it was identified that decision tree model is superior for deriving significant factors and explaining rules. Besides, McNemar Test was conducted to examine difference of errors between the actual value of 4 dependant variables and the classification value from analysis. As a result, with the significance probability of .000, it showed that there is statistically significant difference between classification accuracies for the decision tree and the multinomial logistic regression. 본 연구는 한국고용정보원에서 발표하는 한국 고령화 패널 연구(Korea Longitudinal Study of Ageing) 의 2012년 자료를 이용하여 모수적 가정을 하는 기존의 통계분석의한계점을 보완할 수 있는 데이터마이닝 기법 중 의사결정나무 모형을 구축함으로써 한국 중고령자의 점진적 은퇴 결정요인 및 규칙을 도출하여 은퇴의사결정을 설명하는데연구의 목적을 두었다. 연구의 주요결과는 다음과 같다. 첫째, 최적의 의사결정나무 모형 구축 결과, 현 일자리에 대한 기술적합도를 매우 낮게 인식하고, 구직 시 제시되는 일자리의 적은 수입으로 인한 어려움을 덜 느끼며, 은퇴계획연령이 68.6세 이상 혹은 66.1세 미만인 경우 중고령자들은 점진적 은퇴를 선택하는것으로 나타났다. 둘째, 완전히 성장한 의사결정나무 모형구축 결과, 최적의 의사결정나무에서 도출된매우 낮은 기술적합도, 적은 수입으로 인한 구직시의 어려움, 68.6세 이상 혹은 66.1세미만의 계획된 은퇴연령의 요인 이외에도, 최종학력이 중졸이상이며, 지난해의 가구총소득이 1,100만원 미만인 경우 점진적 은퇴를 선택하는 것으로 나타났다. 반면 기술적합도를 높게 느끼는 경우에는 현 일자리에 대한 높은 만족도, 68.6세 이상의 은퇴계획연령이점진적 은퇴의 선택요인이었다. 이를 통해 그 어떤 요인보다도 중고령자의 직업환경과구직환경, 그리고 계획한 은퇴연령이 주요한 영향요인임을 알 수 있었으며, 특히 은퇴계획연령의 경우 현재 평균 퇴직연령인 52.6세 및 법상에서 명시된 정년연령인 60세와도차이가 나고 있음을 확인 할 수 있었다. 셋째, 구축된 의사결정나무 모형의 우수성을 검증하기 위해 다항 로짓 회귀분석의 결과와 비교해본 결과 의사결정나무의 전체 에러율이 2.30%, 다항 로짓 회귀분석의 전체에러율이 5.40%로 나타나 의사결정나무 모형이 영향요인 도출 및 규칙 설명에 더 우수함을 확인할 수 있었다. 또한 네 가지 종속변수의 실제값과 분석결과로 나타난 분류값간의 오차에 대한 차이를 살펴보기 위해 실시한 맥네머 검정(McNemar Test) 결과에서도 유의확률이 .000으로 의사결정나무와 다항 로짓 회귀분석의 분류 정확도에는 통계적으로 유의미한 차이가 나타나고 있음을 확인하였다.

경영분석지표와 의사결정나무기법을 이용한 유상증자 예측모형 개발

김명균(Myeong-Kyun Kim),조윤호(Yoonho Cho) 한국지능정보시스템학회 2012 지능정보연구 Vol.18 No.4

This study focuses on predicting which firms will increase capital by issuing new stocks in the near future. Many stakeholders, including banks, credit rating agencies and investors, performs a variety of analyses for firms’ growth, profitability, stability, activity, productivity, etc., and regularly report the firms’ financial analysis indices. In the paper, we develop predictive models for rights issues using these financial analysis indices and data mining techniques. This study approaches to building the predictive models from the perspective of two different analyses. The first is the analysis period. We divide the analysis period into before and after the IMF financial crisis, and examine whether there is the difference between the two periods. The second is the prediction time. In order to predict when firms increase capital by issuing new stocks, the prediction time is categorized as one year, two years and three years later. Therefore Total six prediction models are developed and analyzed. In this paper, we employ the decision tree technique to build the prediction models for rights issues. The decision tree is the most widely used prediction method which builds decision trees to label or categorize cases into a set of known classes. In contrast to neural networks, logistic regression and SVM, decision tree techniques are well suited for high-dimensional applications and have strong explanation capabilities. There are well-known decision tree induction algorithms such as CHAID, CART, QUEST, C5.0, etc. Among them, we use C5.0 algorithm which is the most recently developed algorithm and yields performance better than other algorithms. We obtained data for the rights issue and financial analysis from TS2000 of Korea Listed Companies Association. A record of financial analysis data is consisted of 89 variables which include 9 growth indices, 30 profitability indices, 23 stability indices, 6 activity indices and 8 productivity indices. For the model building and test, we used 10,925 financial analysis data of total 658 listed firms. PASW Modeler 13 was used to build C5.0 decision trees for the six prediction models. Total 84 variables among financial analysis data are selected as the input variables of each model, and the rights issue status (issued or not issued) is defined as the output variable. To develop prediction models using C5.0 node (Node Options: Output type = Rule set, Use boosting = false, Cross-validate = false, Mode = Simple, Favor = Generality), we used 60% of data for model building and 40% of data for model test. The results of experimental analysis show that the prediction accuracies of data after the IMF financial crisis (59.04% to 60.43%) are about 10 percent higher than ones before IMF financial crisis (68.78% to 71.41%). These results indicate that since the IMF financial crisis, the reliability of financial analysis indices has increased and the firm intention of rights issue has been more obvious. The experiment results also show that the stability-related indices have a major impact on conducting rights issue in the case of short-term prediction. On the other hand, the long-term prediction of conducting rights issue is affected by financial analysis indices on profitability, stability, activity and productivity. All the prediction models include the industry code as one of significant variables. This means that companies in different types of industries show their different types of patterns for rights issue. We conclude that it is desirable for stakeholders to take into account stability-related indices and more various financial analysis indices for short-term prediction and long-term prediction, respectively. The current study has several limitations. First, we need to compare the differences in accuracy by using different data mining techniques such as neural networks, logistic regression and SVM. Second, we are required to develop and to evaluate new prediction model

A Study on the Classification of Variables Affecting Smartphone Addiction in Decision Tree Environment Using Python Program

김승재 한국인터넷방송통신학회 2022 International journal of advanced smart convergenc Vol.11 No.4

Since the launch of AI, technology development to implement complete and sophisticated AI functions has continued. In efforts to develop technologies for complete automation, Machine Learning techniques and deep learning techniques are mainly used. These techniques deal with supervised learning, unsupervised learning, and reinforcement learning as internal technical elements, and use the Big-data Analysis method again to set the cornerstone for decision-making. In addition, established decision-making is being improved through subsequent repetition and renewal of decision-making standards. In other words, big data analysis, which enables data classification and recognition/recognition, is important enough to be called a key technical element of AI function. Therefore, big data analysis itself is important and requires sophisticated analysis. In this study, among various tools that can analyze big data, we will use a Python program to find out what variables can affect addiction according to smartphone use in a decision tree environment. We the Python program checks whether data classification by decision tree shows the same performance as other tools, and sees if it can give reliability to decision-making about the addictiveness of smartphone use. Through the results of this study, it can be seen that there is no problem in performing big data analysis using any of the various statistical tools such as Python and R when analyzing big data.

상관분석 및 의사결정나무를 통한 사출 성형품 변형의 영향 인자 분석

이용선,이은녕,황순환,한성렬 한국기계기술학회 2022 한국기계기술학회지 Vol.24 No.1

This paper analyzed the correlation between injection molding factors through correlation analysis. In addition, the decision-tree model, which is a white box model with excellent explanatory power, was used to obtain optimal molding conditions that satisfy multiple constraint conditions. First, 243 data to be used in the experiment were created through a full factorial design. Second, a correlation analysis was conducted to understand the correlation. Third, to verify the decision-tree model, the prediction performance was evaluated using RMSE. As a result, good prediction performance was confirmed. A decision-tree experiment analysis was conducted. As a result of the progress, the same results as the correlation analysis were derived. Based on the previous analysis results, optimal molding conditions were applied to CAE. As a result, the amount of deformation in the multi-cavity could be improved by about 1.1% and 2.72% while satisfying the constraint.

Two-Stage Decision Tree Analysis for Diagnosis of Personal Sasang Constitution Medicine Type

진희정,이혜정,김명건,김홍기,김종열,Jin, Hee-Jeong,Lee, Hae-Jung,Kim, Myoung-Geun,Kim, Hong-Gie,Kim, Jong-Yeol The Society Of Sasang Constitutional Medicine 2010 사상체질의학회지 Vol.22 No.3

1. Objectives: In SCM, a personal Sasang constitution must be determined accurately before any Sasang treatment. The purpose of this study is to develop an objective method for classification of Sasang constitution. 2. Methods: We collected samples from 5 centers where SCM is practiced, and applied two-stage decision tree analysis on these samples. We recruited samples from 5 centers. The collected data were from subjects whose response to herbal medicine was confirmed according to Sasang constitution. 3. Results: The two-stage decision tree model shows higher classification power than a simple decision tree model. This study also suggests that gender must be considered in the first stage to improve the accuracy of classification. 4. Conclusions: We identified important factors for classifying Sasang constitutions through two-stage decision tree analysis. The two-stage decision tree model shows higher classification power than a simple decision tree model.

A Comparative Study of Medical Data Classification Methods Based on Decision Tree and System Reconstruction Analysis

Tang, Tzung-I,Zheng, Gang,Huang, Yalou,Shu, Guangfu,Wang, Pengtao Korean Institute of Industrial Engineers 2005 Industrial Engineeering & Management Systems Vol.4 No.1

This paper studies medical data classification methods, comparing decision tree and system reconstruction analysis as applied to heart disease medical data mining. The data we study is collected from patients with coronary heart disease. It has 1,723 records of 71 attributes each. We use the system-reconstruction method to weight it. We use decision tree algorithms, such as induction of decision trees (ID3), classification and regression tree (C4.5), classification and regression tree (CART), Chi-square automatic interaction detector (CHAID), and exhausted CHAID. We use the results to compare the correction rate, leaf number, and tree depth of different decision-tree algorithms. According to the experiments, we know that weighted data can improve the correction rate of coronary heart disease data but has little effect on the tree depth and leaf number.

데이터마이닝을 이용한 보험기업의 CRM 성과 평가에 관한 연구

송만석,박종환,김삼원,조윤재 한국경영컨설팅학회 2008 경영컨설팅연구 Vol.8 No.3

환경의 급격한 변화에 따라 시장경쟁우위 확보와 경영성과의 확보를 위해 전 산업분야에 걸쳐서 CRM에 대한 많은 관심을 가지면서 효율적인 CRM 운영을 위해 CRM 시스템을 경쟁적으로 도입하고 있으나 그 성과에 대한 실효성에 의문이 제기되고 있다. CRM에 대한 기대가 커짐에 따라 CRM과 관련된 많은 연구들이 수행되어져 왔으나, 주로 CRM에 대한 정의와 개념적인 연구 그리고 기업의 CRM성과에 영향을 미치는 인과관계 연구가 이루어져 왔다. 본 연구에서는 데이터마이닝 기법을 이용하여 CRM 시스템을 도입한 보험기업 근무자를 대상으로 설문지를 이용하여 자기평가 기입법으로 작성하도록 하여 CRM성과를 평가하는 것을 수행하였다. 자료처리를 위해 사용된 도구로는 SPSS사의 Answer Tree Version 3.0 통계 패키지를 사용하여 Decision Tree Analysis 분석을 실시하였으며, 사용한 알고리즘은 CHAID 분석으로 유의수준은 .05로 설정하여 실시하였다. 이와 같은 절차 및 자료분석을 통하여 보험기업의 CRM 성과를 평가하였으며, 수익 향상과 시장점유율 상승 그리고 근무 부서별 CRM성과의 시사점을 논의하였다. With the rapid changes in Corporate management environment people began to pay attention to CRM in all the industries to secure the market competitiveness and management achievements. Although they introduce CRM system competitively for more efficient CRM operation, the performance and effectiveness are in doubt. As the expectations on CRM has grown, many researches and studies have been performed regarding CRM. However most of them focused on the definition and conceptual study on CRM or the cause and effects influencing the CRM performance. In this study data mining was used to evaluate CRM performance. The workers in insurance Enterprise which had introduced CRM system were asked to answer the survey using self evaluation. For data processing, SPSS's Answer Tree Version 3.0 statistics package was used for the Decision Tree Analysis. The algorithm used was CHAID analysis and significant level was set as .05.Through these procedures and material analysis, CRM performance of insurance companies was evaluated and the implications of the profit enhancement, the market share increase and CRM performance of each department were discussed.