• 한국지구과학회지
• /
• 제23권2호
• /
• pp.147-153
• /
• 2002

Reasonable and reliable assessment method is one of the most important issues in science education, Partial credits method is an effective tool for assessing students' science inquiry problem solving. The purposes of this study were to classify the Problem solving types based on the analysis of the thinking Process, and how much the related science concept and the science process skills were used in solving science inquiry problems, and to describe the possibility and rationality of the assessment method that gives partial credit 128 high school seniors were selected and their answers were analyzed to identify science concepts they used to solve each problem, and the result was used as the criterion in the scientific concept test development. Also, to study the science inquiry problem solving type, 152 high school seniors were selected, and protocols were made from audio-taped data of their problem solving process through a think-aloud method and retrospective interviews. In order to get a raw data needed in statistical comparison of reliability, discrimination and the difficulty of the test and the production of the regression equation that determines the ratio of partial credit, 640 students were selected and they were given a science inquiry problem test, a science process skills test, and a scientific concept test. Research result suggested it is more reasonable and reliable to switch to the assessment method that applies partial credit to different problem solving types based on the analysis of the thinking process in problem solving process, instead of the dichotomous credit method.

• Asia pacific journal of information systems
• /
• 제21권2호
• /
• pp.43-58
• /
• 2011

Support vector machines (SVMs), a machine learning technique, has been applied to not only binary classification problems such as bankruptcy prediction but also multi-class problems such as corporate credit ratings. However, in general, the performance of SVMs can be easily worse than the best alternative model to SVMs according to the selection of predictors, even though SVMs has the distinguishing feature of successfully classifying and predicting in a lot of dichotomous or multi-class problems. For overcoming the weakness of SVMs, this study has proposed an approach for selecting features for multi-class SVMs that utilize the impurity measures of classification trees. For the selection of the input features, we employed the C4.5 and CART algorithms, including the stepwise method of discriminant analysis, which is a well-known method for selecting features. We have built a multi-class SVMs model for credit rating using the above method and presented experimental results with data regarding S&P 500 companies.

• 지능정보연구
• /
• 제16권4호
• /
• pp.213-225
• /
• 2010

본 연구에서는 기업의 마케팅 프로모션에 따른 반응고객의 구매액 예측을 위한 방법을 제시하고 SVR의 효과적인 학습방법을 제시하였다. 프로모션에 의한 고객의 구매액을 기반으로 고객을 5등급으로 등급화하고 각 등급 내에서 SVR을 적용하여 고객의 구매액을 예측하였다. 본 연구에서 제안하는 예측된 고객의 등급 내에서 고객 구매액을 예측하는 분리데이터 학습법이 프로모션에 반응한 모든 고객을 대상으로 구매액을 예측하는 전체데이터 학습법보다 높은 예측성과를 보여주었다. 일반적으로 세분화된 고객집단을 하나의 집단으로 보고 동일한 마케팅 전략을 제시하나 본 연구를 통해 구매액에 따라 등급화 된 고객의 등급 내에서 다시 고객의 거래 구매액을 예측하여 동일한 집단 내에서도 차별화된 마케팅 전략을 제시할 수 있는 기반을 제시하였다. 즉 동일한 등급에서도 고객 구매액에 따라 고객의 우선순위를 정할 수 있으며, 이는 마케팅 담당자가 프로모션을 제시할 고객을 선정할 때 유용한 정보로 활용될 수 있다.