• Genomics & Informatics
• /
• v.20 no.2
• /
• pp.23.1-23.9
• /
• 2022

A survival prediction model has recently been developed to evaluate the prognosis of resected nonmetastatic pancreatic ductal adenocarcinoma based on a Cox model using two nationwide databases: Surveillance, Epidemiology and End Results (SEER) and Korea Tumor Registry System-Biliary Pancreas (KOTUS-BP). In this study, we applied two machine learning methods-random survival forests (RSF) and support vector machines (SVM)-for survival analysis and compared their prediction performance using the SEER and KOTUS-BP datasets. Three schemes were used for model development and evaluation. First, we utilized data from SEER for model development and used data from KOTUS-BP for external evaluation. Second, these two datasets were swapped by taking data from KOTUS-BP for model development and data from SEER for external evaluation. Finally, we mixed these two datasets half and half and utilized the mixed datasets for model development and validation. We used 9,624 patients from SEER and 3,281 patients from KOTUS-BP to construct a prediction model with seven covariates: age, sex, histologic differentiation, adjuvant treatment, resection margin status, and the American Joint Committee on Cancer 8th edition T-stage and N-stage. Comparing the three schemes, the performance of the Cox model, RSF, and SVM was better when using the mixed datasets than when using the unmixed datasets. When using the mixed datasets, the C-index, 1-year, 2-year, and 3-year time-dependent areas under the curve for the Cox model were 0.644, 0.698, 0.680, and 0.687, respectively. The Cox model performed slightly better than RSF and SVM.

• 전자공학회논문지 IE
• /
• v.44 no.3
• /
• pp.35-41
• /
• 2007

In this paper we suggest a model that the virtual robot based on artificial intelligence performs learning with compensation signals and compare the leaning speed of the virtual robot according to the compensation method after applying it to three type environments. As a result our model has showed that positive compensation is superior to hybrid one mixed positive and negative if there are enough time for learning in case of more or less complicated environment with the numerous foods, obstacles and robots. Otherwise hybrid method is better than positive one.

• Proceedings of the KIEE Conference
• /
• 2006.11a
• /
• pp.192-194
• /
• 2006

In this paper, the situation of energy efficiency standards for motors and diffusion states are analyzed and a new methodology is proposed using diffusion models and learning curves. The existing diffusion models could not explain affects from new appliances' penetration during the diffusion. But a mixed diffusion model with learning curves or learning ratio is studied to explain this penetration.

• Journal of the Korea Society of Computer and Information
• /
• v.21 no.9
• /
• pp.165-176
• /
• 2016

This study developed the learning ability-based smart education program. The effectiveness of the developed materials was investigated using the quantitative-qualitative mixed method, and the process and results of the investigation are as follows. The quantitative investigation was conducted using the non-equivalent pretest-posttest design, in which the smart education method was applied to the experimental group, while the conventional education method was applied to the control group to analyze students' creative problem-solving potential, task concentration, and the variables required for the learning activity. The results showed significantly higher performance in the experimental group over the control group. Regarding data collection in the qualitative investigation, an analysis of the class from the instructor and class consultation logs from the class analyst were collected; the comments on the experience of each class period were collected from students. The results of the analysis of the data suggest that the perception of smart education improved for the instructor, class analyst, and learners as the course progressed.

• Journal of Engineering Education Research
• /
• v.21 no.5
• /
• pp.32-39
• /
• 2018

The purpose of this study was to develop and validate a behaviorally anchored rating scale for peer evaluation in group projects based on social interdependence theory. A mixed method involving a qualitative and quantitative approach was used in this study. In the qualitative study, both the individual and group interviews were conducted to college students regarding their cooperative learning experiences. Data from this qualitative research was analyzed based on 5 elements of cooperative learning and 'critical incidents' were extracted from students' own voices that would serve as specific rating criteria in the scale. Once the 'critical incidents' have been incorporated into the scale, validation from 3 independent experts was conducted. In the quantitative study, correlations with relevant variables were analyzed to examine the criteria-referenced validity. Findings suggest that the behaviorally anchored rating scale for peer evaluation in group projects can be used in various team-based learning contexts.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.10 no.6
• /
• pp.125-130
• /
• 2010

Cooperative learning is a successful teaching strategy in which small teams, each with students of different levels of ability, use a variety of learning activities to improve their understanding of a subject. Each member of a team is responsible not only for learning what is taught but also for helping teammates learn, thus creating an atmosphere of achievement. In this study, we have propose an english, math education program to the children of elementary school and cooperative learning program technique was applied to implement the program. By cooperative learning program, learners will be performed at the same time learning cooperatively. Finally, we have implement a prototype of cooperative learning program and take a usability test with elementary school children. A complementary team to score and mixed was found to be most effective.

• Journal of The Korean Association of Information Education
• /
• v.13 no.4
• /
• pp.497-508
• /
• 2009

As the educational use of intelligent service robots has been proved to be effective, educational service robots have been utilized in kindergarten. In addition, service robots will be used in elementary schools from 2010 for the after-school English program. This trend indicates that r-Learning using service robots will become a major educational paradigm in preparing for future education. This article consists of the following four parts. First, the concept and the type of educational robots were defined and the trend of previous research was examined. Second, the characteristics of robot-assisted learning were analyzed as a part of r-Learning, and difference between r-Learning and u-Learning was compared. Third, the contents and service using a robot-assisted learning system were discussed, the models and trend of service using the robot-assisted learning system were examined, and the aspects of viewing evolution were compared. Finally, suggestions for activating the service market of robot-assisted learning were made for the educational institution, research institution, government and robot companies.

• The Journal of the Korea Contents Association
• /
• v.21 no.11
• /
• pp.135-144
• /
• 2021

Recently, data-driven decision-making technology has become a key technology leading the data industry, and machine learning technology for this requires high-quality training datasets. However, real-world data contains missing values for various reasons, which degrades the performance of prediction models learned from the poor training data. Therefore, in order to build a high-performance model from real-world datasets, many studies on automatically imputing missing values in initial training data have been actively conducted. Many of conventional machine learning-based imputation techniques for handling missing data involve very time-consuming and cumbersome work because they are applied only to numeric type of columns or create individual predictive models for each columns. Therefore, this paper proposes a new data imputation technique called 'Denoising Self-Attention Network (DSAN)', which can be applied to mixed-type dataset containing both numerical and categorical columns. DSAN can learn robust feature expression vectors by combining self-attention and denoising techniques, and can automatically interpolate multiple missing variables in parallel through multi-task learning. To verify the validity of the proposed technique, data imputation experiments has been performed after arbitrarily generating missing values for several mixed-type training data. Then we show the validity of the proposed technique by comparing the performance of the binary classification models trained on imputed data together with the errors between the original and imputed values.

• Journal of the Korean Chemical Society
• /
• v.68 no.2
• /
• pp.93-106
• /
• 2024

This study investigated differences in presence, immersion, and situational interest in small group learning using augmented reality, based on the degree of tool sharing. 84 eighth-grade students participated in small groups of four. Each group was randomly assigned to one of three environments based on marker and device sharing: the shared environment (shared marker and device usage), the mixed environment (shared marker and individual device usage), and the individual environment (individual marker and device usage). Small group learning using augmented reality was conducted for three class periods, focusing on the "Characteristics of Matter" unit. One-way ANOVA results for the dependent variables revealed that, compared to the shared environment, presence and situational interest were significantly higher in the mixed environment, while immersion and situational interest were significantly higher in the individual environment. MANOVA results for the sub-components of each dependent variable showed significant differences in realness for presence, antecedents and experiences for immersion, and instant enjoyment, novelty, and total interest for situational interest. Analysis of interviews and classroom observations indicated that students in shared and individual environments tended to use their devices individually when utilizing augmented reality. However, in mixed environments, students showed a tendency to use their devices collaboratively, leading to more active interactions. Based on these findings, environments for using tools to enhance the effectiveness of small group learning using augmented reality are discussed.

• Journal of the Korea Society of Computer and Information
• /
• v.24 no.11
• /
• pp.41-49
• /
• 2019

We are to build an unsupervised machine learning-based language model which can estimate the amount of information that are in need to process words consisting of subword-level morphemes and syllables. We are then to investigate whether the reading times of words reflecting their morphemic and syllabic structures are predicted by an information-theoretic measure such as surprisal. Specifically, the proposed Morfessor-based unsupervised machine learning model is first to be trained on the large dataset of sentences on Sejong Corpus and is then to be applied to estimate the information-theoretic measure on each word in the test data of Korean words. The reading times of the words in the test data are to be recruited from Korean Lexicon Project (KLP) Database. A comparison between the information-theoretic measures of the words in point and the corresponding reading times by using a linear mixed effect model reveals a reliable correlation between surprisal and reading time. We conclude that surprisal is positively related to the processing effort (i.e. reading time), confirming the surprisal hypothesis.