Earticle

We live in the age of artificial intelligence, and big data and artificial intelligence education are no longer just for majors, but are required to be able to handle non-majors as well. Software and artificial intelligence education for non-majors is not just a general education, it creates talents who can understand and utilize them, and the quality of education is increasingly important. Through such education, we can nurture creative talents who can create and use new values by fusion with various fields of computing technology. Since 2015, many universities have been implementing software-oriented colleges and AI-oriented colleges to foster software-oriented human resources. However, it is not easy to provide AI basic statistics education of big data analysis deception to nonmajors. Therefore, we would like to present a big data education model for non-majors in big data analysis so that big data analysis can be directly applied.

This research aims to analyze the Sasang constitutions of army soldiers using the measuring tools which were developed by the Korea Institute of Oriental Medicine the government-affiliated Ministry of Science. Then, this research compares the The Sasang constitutions of soldiers to the Sasang typology of Yi Je-Ma, who systemized Korean traditional medicine in 1800s. Also, by comparing each result of subjected army unit, we will recommend proper diet management at regular intervals in order to enhance morals of soldiers. Finally, we also suggest to apply Sasang constitution to consultation in military so as to strengthen the combat power of the army.

이 연구는 수학과 과학, 특히 생명과학과의 연결성에 기초하여 미적분과 통계 기본 교과 의 확률의 뜻과 활용 단원을 생명과학Ⅰ의 유전 내용과 통합하여 지도하는 교수-학습 자료 를 개발하고 개발된 교수-학습 자료를 실제 수업에 적용하여 수업에 참여한 고등학생들이 드러내는 특징을 분석함으로써 그 교수학적 시사점을 기술하는데 목적을 두었다. 선행연구 분석을 통해 통합 교수-학습 자료 개발을 위한 세부 관점을 상세화하고 생명과학 교과서와 수학 교과서를 분석하여 연계 가능한 내용을 추출한 다음 이를 실제 교수-학습 자료로 조직 하였다. 이를 활용한 수업에서 학생들이 보이는 특징을 4가지로 요약하여 제시함으로써 수학 과 과학 통합교육의 설계와 실행에 있어 몇 가지 시사점을 제안하였다.

This study developed teaching and learning materials for an integrated education program of probability and genetics in the light of connections between mathematics and biological science. It also analysed characteristics of high school students' mathematical activities which appeared while the students took part in lessons where the developed materials were contributed in order to teach them. To achieve the aim, this study firstly specified five details for the development of the materials based on the results of previous research and extracted contents of probability and genetics which had the possibility of being taught in the integrated education program by examining the text books. After embodying the teaching materials according to the five details and the extracted contents, the researchers implemented 10 lessons by using the materials. This study elaborated some implications for a succeeding integrated education of mathematics and biological science in term of anlaysis results of features from the students' mathematical understanding and attitudes emerging in the lessons.

In this paper we study some basic statistics in quantile regression. In particular, we investigate the residual, goodness-of-fit statistic and the effect of one or few observations on estimates of regression coefficients. In addition, we compare the proposed goodness-of-fit statistic with the statistic considered by Koenker and Machado (1999). An illustrative example based on real data sets is given to see the numerical performance of the proposed basic statistics.

Statistical information about spoken Korean was obtained. The data are the results of analyzing the Korean phonetic dictionary. This is one of the basic survey to make phoneme ballanced corpus of Korean Speech Data Base (KSDB).

본 연구는 협업필터링 알고리즘에서 선호도 예측 전에 선호도 평가 정보를 이용하여 예측 정확도를 향상시킬 수 있는 방법에 대해 연구하였다. 사용자 기반의 방법에서 이웃기반 협업필터링 알고리즘과 아이템 기반에서 평가치 가중합 알고리즘을 선호도 예측에 사용하였다. 이웃기반 협업필터링 알고리즘에서 유사도 가중치로 Pearson 상관계수와 코사인 벡터를 사용하였으며 유사도 가중치를 공통 응답쌍으로 보정해 주는 유의성 가중치를 적용하였다. 아이템 기반의 가중합 알고리즘에는 조정 코사인 벡터를 유사도 가중치로 사용하고 유의성 가중치도 적용하였다. 아이템 기반으로 제안된 가중합 알고리즘을 사용자 기반으로 적용하여 조정 코사인 벡터와 유의성 가중치를 적용한 예측도 실시하였다. 이상 8가지 방법에 의해 예측된 결과를 예측 전 사용자별 선호도 평가치의 평균과 표준편차를 기준으로 4분위로 구분한 후 각 분위에 포함된 사용자들의 예측 결과를 비교하였다. 각 분위에서 예측 결과가 가장 좋은 알고리즘을 선정하고 이들의 조합으로 예측된 결과를 8가지 결과와 비교하였다. 결과에서 평균의 4분위로 구성한 알고리즘 조합의 결과가 가장 우수함을 알 수 있었으며 선호도 예측 이전의 사전 정보를 이용하여 선호도 예측 정확도를 향상시키는 방법이 유효함을 알 수 있었다.

This study is devoted to improve prediction accuracy of collaborative filtering algorithm apply pre-evaluating each user’s ratings before prediction. Neighborhood based collaborative filtering algorithms in user based method and weighted sum algorithms in item based method are applied to preference prediction of user. Each of pearson correlation coefficient and cosine vector as similarity weight is applied to user based collaborative filtering algorithm and significance weight which devalues similarity weight according to numbers of rated items in common is also applied to algorithm. Weighted sum algorithm which is included item based collaborative filtering algorithm uses adjusted cosine vector as similarity weight and also applies significance weight to that algorithm. The results predicted by the above 8 methods are compared to each user’s prediction accuracy which are divided ratings quartile by each user’s mean and standard deviation before prediction. The best result of the 8 prediction methods in each quartile is selected and combine them to compare with the results of each 8 prediction. The results shows that the combination of algorithms, which is composed of each users’ mean quartile, is the most accurate and using pre-evaluation prior to preference prediction is a valid way to improve prediction accuracy.

This thesis reorganized the time series of trade statistics of Busan Port trade during the Japanese colonial period, and analyzed the trade trend based on this. During the Japanese colonial period, the annual average national trade growth rate was very high, around 10%. As a result, Joseon rapidly transformed into an open economy despite being an agricultural society. However, the bias toward Japan in trade was extremely unreasonable, and it was found that the Joseon economy was completely subsumed by the Japanese economy during the Japanese colonial era. In addition, Japan cut off the Joseon economy from other foreign countries thoroughly and subordinated it to serve only the Japanese economy. Busan Port was the one of the key channel of Japanese colonial rule. Thus, the proportion of Busan Port in the total trade of Joseon showed an increasing trend, accounting for around 30%. From 1912, Busan Port overtook Incheon Port to become the nation's largest trading port. As Busan became the largest port in Joseon, the port of Busan was changing from a port more specialized in exporting to Japan to a port in which import from Japan was more important in trade after the annexation. During the Japanese colonial era, the trade balance was chronically in deficit both nationwide and at Busan Port. This deficit was made up by the outflow of gold and silver and the inflow of Japanese capital. In the late 1930s, the export-import balance showed a surplus because goods produced in Joseon and goods imported from Japan increased in export to regions within the en(円) block. Busan Port, which was geographically close to Japan, had a particularly high concentration of trade with Japan. During the Pacific War, the volume of trade with Joseon as a whole decreased based on current prices, but the volume of trade at Busan Port increased. Overall, the trade fluctuations at Busan Port during the Japanese colonial period were more sensitive than nationwide. This uniqueness was due to the geographical proximity to Japan. The largest trading port in Joseon, a much higher share of trade than other ports and high sensitivity to changes in the Japanese economy can be pointed out as major characteristics of Busan Port trade during the Japanese colonial period.

최근 과학기술정보 콘텐츠 분석 기반의 R&D 전략 수립이 중요시됨에 따라 국내 외 과학기술 통계 정보를 수집, 범주화하여 DB를 구축하고 이를 증거 기반의 기초연구정책을 수립하는데 활용하고자 하였다. 국내 외 49종의 과학기술 관련 보고서 등 자료를 수집하여 6,446개 지표 추출, 877개 통계 정보로 범주화, 이중 208개를 집중 관리 통계로 도출하였다. 이렇게 구축된 DB의 활용성 제시를 위해 논문정보 관련 심층분석을 통해 제한적으로나마 우리나라 기초연구의 효과성을 보여주고자 하였다. 이를 위해, 고피인용 논문창출과 R&D 투자 간의 통계기반 분석과 고피인용 논문의 국내 외 지식맵 비교 분석을 실시하였다. 그 결과, 기초연구비 증가가 우리나라의 보편적 연구활동도와 세계적 연구성과 창출에 긍정적 영향을 미친 것으로 나타났다. 또한 순수기초연구분야로 간주되어온 수학, 물리, 화학이 신성장동력이라 불리는 AI, 청정에너지 등 기술 발전의 기반이 되어왔음을 알 수 있었다.

As increasing the importance of the R&D strategy planning based on S&T evidence, the S&T DB was conducted to develop appropriate basic research support strategies by collecting 49 multiple-nations' statistics and information, extracting 6446 raw data, categorizing 877 indicators including 208 core indicators. An statistical and knowledge map analysis using the highly cited publication-related indicators was conducted to examine the expansion of DB utilization by demonstrating effectiveness of Korea's basic research. As a result, basic research investment have a strong influence on creating outstanding R&D outcomes and on producing a foundation of various S&T based-growth engines.

Based on the contents of probability and statistics areas in the seventh curriculum, we developed a statistics education model programmed by visual basic for upper grade students of elementary schools. The model that was consisted of ten modules can be easily manipulated by elementary schools students. The objects with a variety of colors are presented on P.C monitor and dynamic functions are performed by clicking a mouse. According to the survey after using the model, students positively evaluated the contents and characteristics and showed great interest in the model. We will upload the execution files of the model on our homepage. One can download and use them free.