• The Journal of the Acoustical Society of Korea
• /
• v.13 no.2E
• /
• pp.58-67
• /
• 1994

Hidden Markov Model (HMM)-based algorithms have been used successfully in many speech recognition systems, especially large vocabulary systems. Although general purpose processors can be employed for the system, they inevitably suffer from the computational complexity and enormous data. Therefore, it is essential for real-time speech recognition to develop specialized hardware to accelerate the recognition steps. This paper concerns with a real-time implementation of an isolated word recognition system based on HMM. The speech recognition system consists of a host computer (PC), a DSP board, and a prototype Viterbi scoring board. The DSP board extracts feature vectors of speech signal. The Viterbi scoring board has been implemented using three field-programmable gate array chips. It employs a hardware-efficient Viterbi scoring architecture and performs the Viterbi algorithm for HMM-based speech recognition. At the clock rate of 10 MHz, the system can update about 100,000 states within a single frame of 10ms.

• Journal of Industrial Convergence
• /
• v.16 no.4
• /
• pp.33-46
• /
• 2018

Personal credit scoring is an effective tool for banks to properly guide decision profitably on granting loans. Recently, many classification algorithms and models are used in personal credit scoring. Personal credit scoring technology is usually divided into statistical method and non-statistical method. Statistical method includes linear regression, discriminate analysis, logistic regression, and decision tree, etc. Non-statistical method includes linear programming, neural network, genetic algorithm and support vector machine, etc. But for the development of the credit scoring model, there is no consistent conclusion to be drawn regarding which method is the best. In this paper, we will compare the performance of the most common scoring techniques such as logistic regression, neural network, and support vector machines using personal credit data of the financial institution in China. Specifically, we build three models respectively, classify the customers and compare analysis results. According to the results, support vector machine has better performance than logistic regression and neural networks.

• Korean Journal of Adult Nursing
• /
• v.17 no.3
• /
• pp.464-473
• /
• 2005

Purpose: The purpose of this study was to evaluate and compare the predictive ability of three mortality scoring systems; Acute Physiology and Chronic Health Evaluation(APACHE) III, Simplified Acute Physiology Score(SAPS) II, and Mortality Probability Model(MPM) II in discriminating in-hospital mortality for intensive care unit(ICU) patients with spontaneous intracerebral hemorrhage. Methods: Eighty-nine patients admitted to the ICU at a university hospital in Daejeon Korea were recruited for this study. Medical records of the subject were reviewed by a researcher from January 1, 2003 to March 31, 2004, retrospectively. Data were analyzed using SAS 8.1. General characteristic of the subjects were analyzed for frequency and percentage. Results: The results of this study were summarized as follows. The values of the Hosmer-Lemeshow's goodness-of-fit test for the APACHE III, the SAPS II and the MPM II were chi-square H=4.3849 p=0.7345, chi-square H=15.4491 p=0.0307, and chi-square H=0.3356 p=0.8455, respectively. Thus, The calibration of the MPM II found to be the best scoring system, followed by APACHE III. For ROC curve analysis, the areas under the curves of APACHE III, SAPS II, and MPM II were 0.934, 0.918 and 0.813, respectively. Thus, the discrimination of three scoring systems were satisfactory. For two-by-two decision matrices with a decision criterion of 0.5, the correct classification of three scoring systems were good. Conclusion: Both the APACHE III and the MPM II had an excellent power of mortality prediction and discrimination for spontaneous intracerebral hemorrhage patients in ICU.

• Journal of the Korean Data and Information Science Society
• /
• v.24 no.1
• /
• pp.85-94
• /
• 2013

As the financial market becomes larger, the loss increases due to the failure of the credit risk managements from the poor management of the customer information or poor decision-making. Thus, the credit risk management also becomes more important and it is essential to develop a credit scoring model, which is a fundamental tool used to minimize the credit risk. Credit scoring models have been studied and developed only for binary target variables. In this paper, we consider other types of target variables such as ordinal multinomial data or longitudinal binary data and suggest credit scoring models. We then apply our developed models to real data and random data, and investigate their performance through Kolmogorov-Smirnov statistic.

• The Mathematical Education
• /
• v.63 no.2
• /
• pp.187-207
• /
• 2024

In this study, we explored the feasibility of automatically scoring descriptive assessment items using GPT-4 based ChatGPT by comparing and analyzing the scoring results between teachers and GPT-4 based ChatGPT. For this purpose, three descriptive items from the permutation and combination unit for first-year high school students were selected from the KICE (Korea Institute for Curriculum and Evaluation) website. Items 1 and 2 had only one problem-solving strategy, while Item 3 had more than two strategies. Two teachers, each with over eight years of educational experience, graded answers from 204 students and compared these with the results from GPT-4 based ChatGPT. Various techniques such as Few-Shot-CoT, SC, structured, and Iteratively prompts were utilized to construct prompts for scoring, which were then inputted into GPT-4 based ChatGPT for scoring. The scoring results for Items 1 and 2 showed a strong correlation between the teachers' and GPT-4's scoring. For Item 3, which involved multiple problem-solving strategies, the student answers were first classified according to their strategies using prompts inputted into GPT-4 based ChatGPT. Following this classification, scoring prompts tailored to each type were applied and inputted into GPT-4 based ChatGPT for scoring, and these results also showed a strong correlation with the teachers' scoring. Through this, the potential for GPT-4 models utilizing prompt engineering to assist in teachers' scoring was confirmed, and the limitations of this study and directions for future research were presented.

• Journal of the Korea Society of Computer and Information
• /
• v.24 no.1
• /
• pp.57-65
• /
• 2019

In this paper, we propose a Rights Management Information(RMI) expression systems for individual sites are integrated and the performance evaluation is performed to find out an efficient comparing and updating method of RMI through various image feature point search techniques. In addition, we proposed a weighted scoring model for both public domain sites and posts in order to use the most latest RMI based on reliable data. To solve problem that most public domain sites are exposed to copyright infringement by providing inconsistent RMI(Rights Management Information) expression system and non-up-to-date RMI information. The weighted scoring model proposed in this paper makes it possible to use the latest RMI for duplicated images that have been verified through the performance evaluation experiments of SIFT and CNN techniques and to improve the accuracy when applied to search engines. In addition, there is an advantage in providing users with accurate original public domain images and their RMI from the search engine even when some modified public domain images are searched by users.

• Korean Journal of Radiology
• /
• v.22 no.9
• /
• pp.1481-1489
• /
• 2021

Objective: To construct a CT-based Fagotti scoring system by analyzing the correlations between laparoscopic findings and CT features in patients with advanced ovarian cancer. Materials and Methods: This retrospective cohort study included patients diagnosed with stage III/IV ovarian cancer who underwent diagnostic laparoscopy and debulking surgery between January 2010 and June 2018. Two radiologists independently reviewed preoperative CT scans and assessed ten CT features known as predictors of suboptimal cytoreduction. Correlation analysis between ten CT features and seven laparoscopic parameters based on the Fagotti scoring system was performed using Spearman's correlation. Variable selection and model construction were performed by logistic regression with the least absolute shrinkage and selection operator method using a predictive index value (PIV) ≥ 8 as an indicator of suboptimal cytoreduction. The final CT-based scoring system was internally validated using 5-fold cross-validation. Results: A total of 157 patients (median age, 56 years; range, 27-79 years) were evaluated. Among 120 (76.4%) patients with a PIV ≥ 8, 105 patients received neoadjuvant chemotherapy followed by interval debulking surgery, and the optimal cytoreduction rate was 90.5% (95 of 105). Among 37 (23.6%) patients with PIV < 8, 29 patients underwent primary debulking surgery, and the optimal cytoreduction rate was 93.1% (27 of 29). CT features showing significant correlations with PIV ≥ 8 were mesenteric involvement, gastro-transverse mesocolon-splenic space involvement, diaphragmatic involvement, and para-aortic lymphadenopathy. The area under the receiver operating curve of the final model for prediction of PIV ≥ 8 was 0.72 (95% confidence interval: 0.62-0.82). Conclusion: Central tumor burden and upper abdominal spread features on preoperative CT were identified as distinct predictive factors for high PIV on diagnostic laparoscopy. The CT-based PIV prediction model might be useful for patient stratification before cytoreduction surgery for advanced ovarian cancer.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.37 no.1
• /
• pp.175-185
• /
• 2017

As the gas is manufactured, handled and used more often due to the continuous increase of gas, the related facility gets expanded and more complex causing small and big accident which causes economic loss including damage for humans and materials. The gas pipeline, the most common gas facility, has the biggest risk of accidents. Especially in the urban area and densely populated areas, the accident due to the high pressure pipeline may cause even more serious damages. To prevent the accident caused by the buried pipeline, it is required for the relevant authorities to evaluate the damage and risk of the whole pipeline system effectively. A risk is usually defined as a possibility or probability of an undesired event happening, and there is always a risk even when the probability of failure is set low once the pipeline is installed or under operation. It is reported that the accident caused by the failure of the pipeline rarely happens, however, it is important to minimize the rate of accidents by analyzing the reason of failure as it could cause a huge damage of humans and property. Therefore, the paper rated the risk of pipelines with quantitative numbers using the qualitative risk analysis method of the Scoring Model. It is assumed that the result could be effectively used for practical maintenance and management of pipelines securing the safety of the pipes.

• International Journal of Quality Innovation
• /
• v.1 no.1
• /
• pp.75-80
• /
• 2000

Objective evaluation is necessary for the company to know the level of its TQM activity and to improve it. This article proposes self-evaluation model for TQM activity through comparison study of the examination viewpoints for the Deming Prize with criteria of the Malcolm Baldridge National Quality Award. Proposed self-valuation model consists of three evaluation categories i.e. management system, management performance and survey/audit system. Evaluation on these categories is done for process and performance by using scoring method. This self-evaluation model is useful for checking the progress of TQM and make company recognize the strength and weakness of its TQM activity, namely, positioning analysis.

• Speech Sciences
• /
• v.10 no.3
• /
• pp.225-239
• /
• 2003

In this paper, we develop a keyword spotting system using vocabulary-independent speech recognition technique, and investigate several non-keyword modeling and post-processing methods to improve its performance. In order to model non-keyword speech segments, monophone clustering and Gaussian Mixture Model (GMM) are considered. We employ likelihood ratio scoring method for the post-processing schemes to verify the recognition results, and filler models, anti-subword models and N-best decoding results are considered as an alternative hypothesis for likelihood ratio scoring. We also examine different methods to construct anti-subword models. We evaluate the performance of our system on the automatic telephone exchange service task. The results show that GMM-based non-keyword modeling yields better performance than that using monophone clustering. According to the post-processing experiment, the method using anti-keyword model based on Kullback-Leibler distance and N-best decoding method show better performance than other methods, and we could reduce more than 50% of keyword recognition errors with keyword rejection rate of 5%.