• Proceedings of the Korea Inteligent Information System Society Conference
• /
• 2003.05a
• /
• pp.474-483
• /
• 2003

A number of studies for corporate bond rating classification problems have demonstrated that artificial intelligence approaches such as Case-based reasoning (CBR) can be alternative methodologies to statistical techniques. CBR is a problem solving technique in that the case specific knowledge of past experience is utilized to find a most similar solution to the new problems. To build a successful CBR system to deal with human information processing, the representation of knowledge of each attribute is an important key factor We propose a hybrid approach of using fuzzy sets that describe the approximate phenomena of the real world because it handles inexact knowledge represented by common linguistic terms in a similar way as human reasoning compared to the other existing techniques. Integration of fuzzy sets with CBR is important to develop effective methods for dealing with vague and incomplete knowledge to statistical represent using membership value of fuzzy sets in CBR.

• KIPS Transactions on Computer and Communication Systems
• /
• v.9 no.1
• /
• pp.17-24
• /
• 2020

In this paper, we present a movie rating prediction architecture based on a convolutional neural network (CNN). Our prediction architecture extends TextCNN, a popular CNN-based architecture for sentence classification, in three aspects. First, character embeddings are utilized to cover many variants of words since reviews are short and not well-written linguistically. Second, the attention mechanism (i.e., squeeze-and-excitation) is adopted to focus on important features. Third, a scoring function is proposed to convert the output of an activation function to a review score in a certain range (1-10). We evaluated our prediction architecture on a movie review dataset and achieved a low MSE (e.g., 3.3841) compared with an existing method. It showed the superiority of our movie rating prediction architecture.

• Geotechnical Engineering
• /
• v.14 no.1
• /
• pp.29-36
• /
• 1998

The application of a multiple rock classification method, which is a generalization of a binary rock classification, is studied in this paper. In particular, this paper shows how to incorporate qualitative data through a case study. The method suggested in this paper can be effectively used for a systematic multiple rock classification such as RMR system developed by Bieniawski. It will be very useful for rock classifications. In addition, it is known that the expected cost of errors can be atopted to indicate how well a investigation plan is made.

• Journal of Digital Convergence
• /
• v.14 no.8
• /
• pp.163-175
• /
• 2016

The purpose of this study is to provide an efficient internal control system formation incentives for company and to confirm empirically usefulness of the internal accounting control system for financial institutions by analyzing whether the internal control vulnerabilities of companies related significantly to the classification and assessment of soundness of financial institutions. Empirical analysis covered KOSPI, KOSDAQ listed companies and unlisted companies with more than 100 billion won of assets which have trading performance with "K" financial institution from 2008 until 2013. Whereas non-internal control vulnerability reporting companies by the internal control of financial reporting received average credit rating of BBB on average, reporting companies received CCC rating. And statistically significantly, non-reporting companies are classified as "normal" and reporting companies are classified as "precautionary loan" when it comes to asset quality classification rating. Therefore, reported information of internal control vulnerability reduced the credibility of the financial data, which causes low credit ratings for companies and suggests financial institutions save additional allowance for asset insolvency prevention and require high interest rates. It is a major contribution of this study that vulnerability reporting of internal control in accordance with the internal control of financial reporting can be used as information significant for the evaluation of financial institutions on corporate soundness.

• Journal of the Korean Society of Clothing and Textiles
• /
• v.22 no.2
• /
• pp.241-249
• /
• 1998

Somatotypes in the lower torso and legs of adult females were classified and discriminated. Then their distribution according to the age groups was studied. The subjects were 343 females of 18 to 49 year-old. 36 anthropometric and 21 photographic measurements were taken to each subject. The results were as follows: 1. The somatotypes in the lower torso and legs were classified into 4 types and their differences were shown by analysing factor scores and anthropometric values according to each somatotype. 2. The shape characteristic of tile classified somatotypes was represented by the rating scale of Heath-Carter. 3. The lateral silhouettes of 4 types were compared with balanced type which is type 3 in this study. 4. Individual somatotype in the lower torso and legs could be discriminated from the measured anthropometric data without modifying the data. Anthropometric data, which are needed for discriminating individual somatotype, are waist circumference, posterior waist height, and hip circumference. 5. The distribution of the somatotypes in each age group showed that the dominant somatotype of each age group was different and any somatotype was shown in a specified age group but rarely in other age group.

• The Journal of Engineering Geology
• /
• v.15 no.4 s.42
• /
• pp.463-473
• /
• 2005

Because of using the same rating ranges for every rock types the RMR or the Q-system could not usually consider local geological characteristics They also could not present sufficiently the engineering anisotropy of rocks. The canonical correlation analysis was carried out with 3 kinds of face mapping data obtained from granite, sedimentary rock and phyllite in order to clarify a discrepancy between rock types. According to analysis results, as a type of rocks changes, RM factors have different influences on the total rating of RMR.

• Proceedings of the Korea Inteligent Information System Society Conference
• /
• 2005.11a
• /
• pp.509-515
• /
• 2005

Recently, support vector machines (SVMs) are being recognized as competitive tools as compared with other data mining techniques for solving pattern recognition or classification decision problems. Furthermore, many researches, in particular, have proved it more powerful than traditional artificial neural networks (ANNs)(Amendolia et al., 2003; Huang et al., 2004, Huang et al., 2005; Tay and Cao, 2001; Min and Lee, 2005; Shin et al, 2005; Kim, 2003). The classification decision, such as a binary or multi-class decision problem, used by any classifier, i.e. data mining techniques is cost-sensitive. Therefore, it is necessary to convert the output of the classifier into well-calibrated posterior probabilities. However, SVMs basically do not provide such probabilities. So it required to use any method to create probabilities (Platt, 1999; Drish, 2001). This study applies a method to estimate the probability of outputs of SVM to bankruptcy prediction and then suggests credit scoring methods using the estimated probability for bank's loan decision making.

• Journal of the Korean Academy of Child and Adolescent Psychiatry
• /
• v.18 no.1
• /
• pp.38-48
• /
• 2007

Objectives : The K-ARS (Korean ADHD Rating Scale) is one of the most important assessment tool of attention-deficit hyperactivity disorder (ADHD) in Korea. in this study, we presented detailed normative data on the K-ARS for school-aged children in Seoul metropolitan area to put it to practical use. Methods : The subjects were 2,397 students(1,223 boys and 1,174 girls, aged 6-12) from 4 elementary schools in Seoul, and one caretaker of each child completed the K-ARS for parents. Children who showed high scores of the K-ARS for parents were screened, and 2 child psychiatrists interviewed them to make a clinical diagnosis. We compared the mean scores of the K-ARS for parents between ADHD and normal group, and examined the percentage of correct classification. Results : There were some differences in score of the K-ARS for parents according to sex and age, so we presented continuous normative data with T score and subdivided cut-off points for ADHD screening. Interviews with child psychiatrists using DSM-IV criteria were performed to test diagnostic validity, and the difference in every the K-ARS for parents index between ADHD and normal group was significant(p<.001). Using 3 different cut-of points(80th, 90th, 93rd percentage), the accuracies of ADHD correct classification were 67.9, 72.2, 71.1% and all 3 canonical discriminants were significant (p<.05) between ADHD and normal group. Conclusion : The normative data and cut-off points on the K-ARS for parents are useful in screening ADHD children in Seoul metropolitan area.

• KIISE Transactions on Computing Practices
• /
• v.21 no.2
• /
• pp.121-125
• /
• 2015

In recent years, research on collaborative filtering-based recommendation systems emphasized the accuracy of rating predictions, and this has led to an increase in computation time. As a result, such systems have divergeded from the original purpose of making quick recommendations. In this paper, we propose a recommendation algorithm that uses a Preference Difference Metric to reduce the computation time and to maintain adequate performance. The system recommends items according to their preference classification.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2023.11a
• /
• pp.628-631
• /
• 2023

We present a new approach to evaluate the generated texts by Large Language Models (LLMs) for meme classification. Analyzing an image with embedded texts, i.e. meme, is challenging, even for existing state-of-the-art computer vision models. By leveraging large image-to-text models, we can extract image descriptions that can be used in other tasks, such as classification. In our methodology, we first generate image captions using BLIP-2 models. Using these captions, we use GPT-4 to evaluate the relationship between the caption and the meme text. The results show that OPT_6.7B provides a better rating than other LLMs, suggesting that the proposed method has a potential for meme classification.