• 산업경영시스템학회지
• /
• 제44권2호
• /
• pp.15-23
• /
• 2021

Most of the open-source decision tree algorithms are based on three splitting criteria (Entropy, Gini Index, and Gain Ratio). Therefore, the advantages and disadvantages of these three popular algorithms need to be studied more thoroughly. Comparisons of the three algorithms were mainly performed with respect to the predictive performance. In this work, we conducted a comparative experiment on the splitting criteria of three decision trees, focusing on their interpretability. Depth, homogeneity, coverage, lift, and stability were used as indicators for measuring interpretability. To measure the stability of decision trees, we present a measure of the stability of the root node and the stability of the dominating rules based on a measure of the similarity of trees. Based on 10 data collected from UCI and Kaggle, we compare the interpretability of DT (Decision Tree) algorithms based on three splitting criteria. The results show that the GR (Gain Ratio) branch-based DT algorithm performs well in terms of lift and homogeneity, while the GINI (Gini Index) and ENT (Entropy) branch-based DT algorithms performs well in terms of coverage. With respect to stability, considering both the similarity of the dominating rule or the similarity of the root node, the DT algorithm according to the ENT splitting criterion shows the best results.

• 대한핵의학회지
• /
• 제3권1호
• /
• pp.83-99
• /
• 1969

There have been reported numberous cases of liver scanning in use of $^{198}Au$ colloid by many investigators, however, one in use of $^{113m}In$ colloid has not been reported as yet in this country. The dose of $^{113m}In$ for high diagnostic value in examination of each organ was determined and the dignostic interpretability of liver scanning with the use of $^{113m}In$ was carefully evaluated in comparison with the results of the liver scanning by the conventionally applied radioisotopes. The comparative study of both figures of liver scannings with the use of $^{113m}In$ colloid and $^{198}Au$ colloid delivered following results: 1. The liver uptake rate and clearance into peripheral blood were accentuated more in case of $^{113m}In$ colloid than in case of $^{198}Au$ colloid. 2. The interpretability of space occupying lesion in liver scanning with $^{113m}In$ was also superior to one with $^{198}Au$. 3. The figure of liver scanning with $^{113m}In$ colloid corresponds not always to the figure with $^{198}Au$. This difference can be explained by differences of phagocytic ability of reticuloendotherial system within liver. 4. In the liver scanning with $^{113m}In$ colloid, the spleen is also visualized even in normal examinee. 5. In the cases of disturbed liver function, uptake is more decreased in use of $^{113m}In$ colloid than in $^{198}Au$, in the spleen, however, the way is contrary. 6. With use of $^{113m}In$ colloid, the time required for scanning could be shortened in comparison with $^{198}Au$. 7. The filtration of $^{113m}In$ colloid for scanning prior to human administration gives an expectation for better scanning figure.

• 제어로봇시스템학회논문지
• /
• 제17권8호
• /
• pp.731-738
• /
• 2011

In this paper, we introduce an advanced architecture of K-Means clustering-based polynomial Radial Basis Function Neural Networks (p-RBFNNs) designed with the aid of SSOA (Space Search Optimization Algorithm) and develop a comprehensive design methodology supporting their construction. In order to design the optimized p-RBFNNs, a center value of each receptive field is determined by running the K-Means clustering algorithm and then the center value and the width of the corresponding receptive field are optimized through SSOA. The connections (weights) of the proposed p-RBFNNs are of functional character and are realized by considering three types of polynomials. In addition, a WLSE (Weighted Least Square Estimation) is used to estimate the coefficients of polynomials (serving as functional connections of the network) of each node from output node. Therefore, a local learning capability and an interpretability of the proposed model are improved. The proposed model is illustrated with the use of nonlinear function, NOx called Machine Learning dataset. A comparative analysis reveals that the proposed model exhibits higher accuracy and superb predictive capability in comparison to some previous models available in the literature.

• 응용통계연구
• /
• 제26권1호
• /
• pp.131-139
• /
• 2013

지지벡터기계는 잡음변수가 존재하는 경우에 성능이 저하될 수 있다. 또한 최종 분류기에서 각 변수들의 중요도를 알리 어려운 단점이 있다. 따라서 변수선택은 지지벡터기계의 해석력과 정확도를 높일 수 있다. 기존의 문헌상의 대부분의 연구는 선형 지지벡터기계에서 성근 해를 주는 벌점함수를 통해 변수를 선택에 관한 것이다. 실제로는 분류의 정확도를 높이기 위해 비선형 커널을 사용하는 경우가 일반적이다. 따라서 변수선택은 비선형 지지벡터기계에서도 마찬가지로 필요하다. 본 논문에서는 모의실험 및 실제자료를 통하여 비선형 지지벡터의 대표적인 변수선택법인 COSSO(component selection and smoothing operator)와 KNIFE(kernel iterative feature extraction)의 성능을 비교한다.

• 한국경영과학회지
• /
• 제38권4호
• /
• pp.95-112
• /
• 2013

Information system (IS) security continues to present a challenge for firms. Especially, IT security accident is recently taking place successively in the financial sector. Thus, a comprehensive measure on this is demanded. A large part of a research on security relies upon technical design in nature and is restrictive in a consideration of person and organizational issue. To achieve a goal of firm security, it is possible with an effort of organizational management and supervision for maintaining the technical and procedural status. Based on a theory of accountability, we propose that the security countermeasures of organization lead to an increase in accountability and reduction in risk of IT security in a financial firm and further to firm performance like promotion in firm reliability. In addition, we investigate which difference a theoretical model shows by comparison between South Korean and American financial firms. As a result of analysis, it found that South Korea and America have significant difference, but that a measure on the financing IT security is important for both countries. We aim to enhance interpretability of a research on security by comparatively analysis between countries and conducting a study focus on specific firm called financial business. Our study suggest new theoretical framework to a research of security and provide guideline on design of security to financial firm.

• 한국농림기상학회지
• /
• 제1권2호
• /
• pp.142-150
• /
• 1999

결론적으로 위성에 의한 강우관측은 지금은 지구의 거의 대부분 지역에서 관측 및 모니터링과 더불어 강우를 이해하는 능력에 커다란 기여를 하고 있다. 특히 1997년 11월 TRMM의 등장으로 강우의 연직분포 및 정량적 관측이 이루어지므로서 국내에서도 위성강우의 활용연구가 급진전할 것으로 기대된다. 또한 위성강우자료를 이용한 농업 및 수자원분야에서의 연구도 기대된다. 위성강우자료를 농업/수자원 분야에 적용하기 위해서는 먼저 자료간의 결합시에 발생되는 규모(scale)와 해상도(resolution) 문제를 다루는 평가연구가 요구된다. 규모문제는 주어진 연구에 가장 적절한 구역의 개수에 대한 불확실성 때문에 발생되며, 해상도 문제는 해당자료를 저해상으로 결합하여 분석하게 되면 기대치 이상의 정보손실을 야기시키기 때문이다. 이와 같은 의문점이 어느 정도 구명된다면 위성강우자료를 이용한 연계연구는 무궁무진하다고 말할 수 있다. 예를 들면 농업분야에서는 위성강우에 의한 격자기반(grid-based)의 토양수분 및 지하수위 변화, 농경지 침수지역 예측 등에, 수자원분야에서는 공간강우-유출해석에 의한 홍수예 · 경보시스템의 향상, 도시지역 홍수범람지역의 예측 등에 활용할 수 있다.

• 전기학회논문지
• /
• 제60권1호
• /
• pp.184-192
• /
• 2011

In this study, we introduce a identification methodology for FCM-based fuzzy model. The two underlying design mechanisms of such networks involve Fuzzy C-Means (FCM) clustering method and Particle Swarm Optimization(PSO). The proposed algorithm is based on FCM clustering method for efficient processing of data and the optimization of model was carried out using PSO. The premise part of fuzzy rules does not construct as any fixed membership functions such as triangular, gaussian, ellipsoidal because we build up the premise part of fuzzy rules using FCM. As a result, the proposed model can lead to the compact architecture of network. In this study, as the consequence part of fuzzy rules, we are able to use four types of polynomials such as simplified, linear, quadratic, modified quadratic. In addition, a Weighted Least Square Estimation to estimate the coefficients of polynomials, which are the consequent parts of fuzzy model, can decouple each fuzzy rule from the other fuzzy rules. Therefore, a local learning capability and an interpretability of the proposed fuzzy model are improved. Also, the parameters of the proposed fuzzy model such as a fuzzification coefficient of FCM clustering, the number of clusters of FCM clustering, and the polynomial type of the consequent part of fuzzy rules are adjusted using PSO. The proposed model is illustrated with the use of Automobile Miles per Gallon(MPG) and Boston housing called Machine Learning dataset. A comparative analysis reveals that the proposed FCM-based fuzzy model exhibits higher accuracy and superb predictive capability in comparison to some previous models available in the literature.

• 한국빅데이터학회지
• /
• 제3권2호
• /
• pp.101-112
• /
• 2018

본 연구는 기계학습의 하나인 신경망 분석과 음이항 회귀분석을 활용하여 경찰신고건수를 예측하고자 2016년 6월부터 2017년 5월까지 충남지방경찰청에 접수된 112신고 데이터를 이용하여 예측모델을 개발하였다. 모델을 개발하기 위해 경찰신고건수에 영향을 줄 수 있는 시간, 휴일, 휴일 전날, 계절, 기온, 강수량, 풍속, 관할면적, 인구, 외국인 수, 단독주택비율, 기타주택비율 변수 등을 활용하였다. 변수의 종류에 따라 몇몇은 경찰신고건수와 양의 상관관계 또는 음의 상관관계가 확인되었다. 사용된 두 개의 방법론을 비교한바, 신경망분석의 예측 결과는 예측 값과 실제 값의 상관계수 0.7702, RMSE 2.557이고, 음이항 회귀분석은 상관계수 0.7158, RMSE 2.831으로 나타났다. 신경망분석은 해석가능성은 낮지만, 음이항 회귀분석에 비해 예측력이 뛰어나다는 것이 확인되었다. 향후 경찰관서에서 본 연구의 예측모델을 기초로 하여 최적의 경찰력 배치를 할 수 있을 것으로 기대된다.