• The Korean Journal of Applied Statistics
• /
• v.36 no.1
• /
• pp.33-48
• /
• 2023

In this paper we consider automatic speech recognition (ASR) for Korean speech data in which elderly persons randomly speak a sequence of words such as animals and vegetables for one minute. Most of the speakers are over 60 years old and some of them are dementia patients. The goal is to compare deep-learning based ASR models for such data and to find models with good performance. ASR is a technology that can recognize spoken words and convert them into written text by computers. Recently, many deep-learning models with good performance have been developed for ASR. Training data for such models are mostly composed of the form of sentences. Furthermore, the speakers in the data should be able to pronounce accurately in most cases. However, in our data, most of the speakers are over the age of 60 and often have incorrect pronunciation. Also, it is Korean speech data in which speakers randomly say series of words, not sentences, for one minute. Therefore, pre-trained models based on typical training data may not be suitable for our data, and hence we train deep-learning based ASR models from scratch using our data. We also apply some data augmentation methods due to small data size.

• KIPS Transactions on Software and Data Engineering
• /
• v.12 no.4
• /
• pp.179-188
• /
• 2023

Recently, fake news disguises the form of news content and appears whenever important events occur, causing social confusion. Accordingly, artificial intelligence technology is used as a research to detect fake news. Fake news detection approaches such as automatically recognizing and blocking fake news through natural language processing or detecting social media influencer accounts that spread false information by combining with network causal inference could be implemented through deep learning. However, fake news detection is classified as a difficult problem to solve among many natural language processing fields. Due to the variety of forms and expressions of fake news, the difficulty of feature extraction is high, and there are various limitations, such as that one feature may have different meanings depending on the category to which the news belongs. In this paper, emotional change patterns are presented as an additional identification criterion for detecting fake news. We propose a model with improved performance by applying a convolutional neural network to a fake news data set to perform analysis based on content characteristics and additionally analyze emotional change patterns. Sentimental polarity is calculated for the sentences constituting the news and the result value dependent on the sentence order can be obtained by applying long-term and short-term memory. This is defined as a pattern of emotional change and combined with the content characteristics of news to be used as an independent variable in the proposed model for fake news detection. We train the proposed model and comparison model by deep learning and conduct an experiment using a fake news data set to confirm that emotion change patterns can improve fake news detection performance.

• The Journal of the Acoustical Society of Korea
• /
• v.42 no.2
• /
• pp.102-111
• /
• 2023

Automatic Music Transcription (AMT) is a task that detects and recognizes musical note events from a given audio recording. In this paper, we focus on reducing the latency of real-time AMT systems on piano music. Although neural AMT models have been adapted for real-time piano transcription, they suffer from high latency, which hinders their usefulness in interactive scenarios. To tackle this issue, we explore several techniques for reducing the intrinsic latency of a neural network for piano transcription, including reducing window and hop sizes of Fast Fourier Transformation (FFT), modifying convolutional layer's kernel size, and shifting the label in the time-axis to train the model to predict onset earlier. Our experiments demonstrate that combining these approaches can lower latency while maintaining high transcription accuracy. Specifically, our modified model achieved note F1 scores of 92.67 % and 90.51 % with latencies of 96 ms and 64 ms, respectively, compared to the baseline model's note F1 score of 93.43 % with a latency of 160 ms. This methodology has potential for training AMT models for various interactive scenarios, including providing real-time feedback for piano education.

• Journal of the Korean Geotechnical Society
• /
• v.24 no.5
• /
• pp.129-137
• /
• 2008

An active application of concrete track is being expected far the future constructions of Korean railroad. In comparison with the existing ballasted tract, a concrete track is very susceptible for the settlement, since its rehabilitation requires much time and cost. When a concrete track is constructed on fine-grained subgrade soil, excessive cumulative plastic settlements due to repetitive train road may occur. In this case, the settlement of the concrete track may be effectively reduced by installing a small number of small-diameter piles beneath the track. This paper presents the effect of pile installation on the reduction of cumulative plastic settlement of concrete track. A method combining experiential equation and numerical method is proposed. Using an existing experiential equation and the estimated earth pressure distribution, the cumulative plastic strain was calculated. From the results, it is verified that the effects of the pile installation is significant to effectively reduce the cumulative plastic settlement of concrete track. The reduction effects of the cumulative plastic settlement according to the pile number and pile arrangement are presented.

• Journal of Industrial Convergence
• /
• v.21 no.5
• /
• pp.109-122
• /
• 2023

The purpose of this study is to look into how elderly people's health promoting behaviors influence their successful aging, to realize how their perception of successful aging and their life satisfaction have the mediating effect on the path from health promotion behaviors to successful aging, and to find the significant influence of successful aging perception and life satisfaction on successful aging. This researcher conducted a questionnaire survey with elderly people using a senior welfare center in Gyeonggio-do, and analyzed 250 copies that. For data analysis, SPSS Win 25 was applied to conduct frequency analysis, descriptive statistics, t-test, one-way ANOVA, and correlation analysis. Mediating effect analysis was made to verify the causal relations between health promoting behaviors and successful aging, and to validate the mediating effect of successful aging perception and life satisfaction on the causal relations. As a result, elderly people's health promoting behaviors influenced their perception of successful aging, their life satisfaction, and their successful aging. Their perception of successful aging had the mediating effect on health promotional behaviors and successful aging, but life satisfaction did not so. This study has the following implications: it is necessary to train persons specializing in support for health promoting, to develop an efficient health promotional model and program, and to provide an opportunity of education. It is necessary to come up with a support policy in consideration of tangible or intangible factors. It is necessary to establish a policy in line with economic levels and health conditions of elderly people.

• Journal of the Korea Society of Computer and Information
• /
• v.14 no.9
• /
• pp.11-19
• /
• 2009

A RAM-based Neural Net(RBNN) which has multi-discriminators is more effective than RBNN with a discriminator. Experience Sensitive Cumulative Neural Network and 3-D Neuro System(3DNS) that accumulate the features point improved the performance of BNN, which were enabled to train additional and repeated patterns and extract a generalized pattern. In recognition process of Neural Net with multi-discriminator, the selection of class was decided by the value of MRD which calculates the accumulated sum of each class. But they had a saturation problem of its memory cells caused by learning volume increment. Therefore, the decision of MRD has a low performance because recognition rate is decreased by saturation. In this paper, we propose the method which improve the MRD ability. The method consists of the optimum MRD and the matching ratio prototype to generalized image, the cumulative filter ratio, the gap of prototype response MRD. We experimented the performance using NIST database of NIST without preprocessor, and compared this model with 3DNS. The proposed MRD method has more performance of recognition rate and more stable system for distortion of input pattern than 3DNS.

• Journal of Popular Narrative
• /
• v.25 no.1
• /
• pp.315-347
• /
• 2019

(1986) and (1987) represent the society of 1980s in which the professional baseball game was initiated to cover the irrational military culture. The love and marriage of sports players were the headlines of the media, and the yearly salary of the players was the hottest issue of conversation. The military culture is represented in the scenes where the coaches train the failures and inapt players in extreme drills. The films pinpoint the absurdity of military culture and win-at-all-costs mentality. The collapse of the dictatorial leadership at the end of the films is a metaphor for the collapse of the fifth Republic of Korea. The episodes where the players talk about contract money, and the trade of players and sports business were a new phenomenon of the 1980's. The fact that Oh Hyesung of chooses love instead of victory deals a big blow to the secular ambition for money, victory and dictatorial leadership. His option provides catharsis for an audience oppressed under military leadership and success driven ideology. On the other hand, Oh Hyesung of dies right at the moment of winning the world champion. He achieves neither love nor success. While Oh Hyesung of is a symbol of pure love and gives spiritual comfort to the audience, Oh Hyesung of gives a sense of hopelessness to the audience. Both of the two sports films reflect the representation of the 1980's but received opposing reviews from audiences.

• Journal of Intelligence and Information Systems
• /
• v.29 no.2
• /
• pp.241-265
• /
• 2023

A corporate insolvency prediction model serves as a vital tool for objectively monitoring the financial condition of companies. It enables timely warnings, facilitates responsive actions, and supports the formulation of effective management strategies to mitigate bankruptcy risks and enhance performance. Investors and financial institutions utilize default prediction models to minimize financial losses. As the interest in utilizing artificial intelligence (AI) technology for corporate insolvency prediction grows, extensive research has been conducted in this domain. However, there is an increasing demand for explainable AI models in corporate insolvency prediction, emphasizing interpretability and reliability. The SHAP (SHapley Additive exPlanations) technique has gained significant popularity and has demonstrated strong performance in various applications. Nonetheless, it has limitations such as computational cost, processing time, and scalability concerns based on the number of variables. This study introduces a novel approach to variable selection that reduces the number of variables by averaging SHAP values from bootstrapped data subsets instead of using the entire dataset. This technique aims to improve computational efficiency while maintaining excellent predictive performance. To obtain classification results, we aim to train random forest, XGBoost, and C5.0 models using carefully selected variables with high interpretability. The classification accuracy of the ensemble model, generated through soft voting as the goal of high-performance model design, is compared with the individual models. The study leverages data from 1,698 Korean light industrial companies and employs bootstrapping to create distinct data groups. Logistic Regression is employed to calculate SHAP values for each data group, and their averages are computed to derive the final SHAP values. The proposed model enhances interpretability and aims to achieve superior predictive performance.

• Korean Chemical Engineering Research
• /
• v.61 no.3
• /
• pp.388-393
• /
• 2023

The analysis of engineering data using neural network based on supervised learning has been utilized in various engineering fields such as optimization of chemical engineering process, concentration prediction of particulate matter pollution, prediction of thermodynamic phase equilibria, and prediction of physical properties for transport phenomena system. The supervised learning requires training data, and the performance of the supervised learning is affected by the composition and the configurations of the given training data. Among the frequently observed engineering data, the data is given in log-scale such as length of DNA, concentration of analytes, etc. In this study, for widely distributed log-scaled training data of virtual 100×100 images, available loss functions were quantitatively evaluated in terms of (i) confusion matrix, (ii) maximum relative error and (iii) mean relative error. As a result, the loss functions of mean-absolute-percentage-error and mean-squared-logarithmic-error were the optimal functions for the log-scaled training data. Furthermore, we figured out that uniformly selected training data lead to the best prediction performance. The optimal loss functions and method for how to compose training data studied in this work would be applied to engineering problems such as evaluating DNA length, analyzing biomolecules, predicting concentration of colloidal suspension.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2020.06a
• /
• pp.261-261
• /
• 2020

물관리의 기본이 되는 연속적인 유량 자료 확보를 위해서는 정확도 높은 수위-유량 관계 곡선식 개발이 필수적이다. 수위-유량 관계곡선식은 모든 수문시설 설계의 기초가 되며 홍수, 가뭄 등 물재해 대응을 위해서도 중요한 의미를 가지고 있다. 그러나 일반적으로 유량 측정은 많은 비용과 시간이 들고, 식생성장, 단면변화 등의 통제특성(control)이 변함에 따라 구간분리, 기간분리와 같은 비선형적인 양상이 나타나 자료 해석에 어려움이 존재한다. 특히, 국내 하천의 경우 자연적 및 인위적인 환경 변화가 다양하여 지점 및 기간에 따라 세밀한 분석이 요구된다. 머신러닝(Machine Learning)이란 데이터를 통해 컴퓨터가 스스로 학습하여 모델을 구축하고 성능을 향상시키는 일련의 과정을 뜻한다. 기존의 수위-유량 관계곡선식은 개발자의 판단에 의해 데이터의 종류와 기간 등을 설정하여 회귀식의 파라미터를 산출한다면, 머신러닝은 유효한 전체 데이터를 이용해 스스로 학습하여 자료 간 상관성을 찾아내 모델을 구축하고 성능을 지속적으로 향상 시킬 수 있다. 머신러닝은 충분한 수문자료가 확보되었다는 전제 하에 복잡하고 가변적인 수자원 환경을 반영하여 유량 추정의 정확도를 지속적으로 향상시킬 수 있다는 이점을 가지고 있다. 본 연구는 머신러닝의 대표적인 알고리즘들을 활용하여 유량을 추정하는 모델을 구축하고 성능을 비교·분석하였다. 대상지역은 안정적인 수량을 확보하고 있는 한강수계의 거운교 지점이며, 사용자료는 2010~2018년의 시간, 수위, 유량, 수면폭 등 이다. 프로그램은 파이썬을 기반으로 한 머신러닝 라이브러리인 사이킷런(sklearn)을 사용하였고 알고리즘은 랜덤포레스트 회귀, 의사결정트리, KNN(K-Nearest Neighbor), rgboost을 적용하였다. 학습(train) 데이터는 입력자료 종류별로 조합하여 6개의 세트로 구분하여 모델을 구축하였고, 이를 적용해 검증(test) 데이터를 RMSE(Roog Mean Square Error)로 평가하였다. 그 결과 모델 및 입력 자료의 조합에 따라 3.67~171.46로 다소 넓은 범위의 값이 도출되었다. 그 중 가장 우수한 유형은 수위, 연도, 수면폭 3개의 입력자료를 조합하여 랜덤포레스트 회귀 모델에 적용한 경우이다. 비교를 위해 동일한 검증 데이터를 한국수문조사연보(2018년) 내거운교 지점의 수위별 수위-유량 곡선식을 이용해 유량을 추정한 결과 RMSE가 3.76이 산출되어, 머신러닝이 세분화된 수위-유량 곡선식과 비슷한 수준까지 성능을 내는 것으로 확인되었다. 본 연구는 양질의 유량자료 생산을 위해 기 구축된 수문자료를 기반으로 머신러닝 기법의 적용 가능성을 검토한 기초 연구로써, 국내 효율적인 수문자료 측정 및 수위-유량 곡선 산출에 도움이 될 수 있을 것으로 판단된다. 향후 수자원 환경 및 통제특성에 영향을 미치는 다양한 영향변수를 파악하기 위해 기상자료, 취수량 등의 입력 자료를 적용할 필요가 있으며, 머신러닝 내 비지도학습인 딥러닝과 같은 보다 정교한 모델에 대한 추가적인 연구도 수행되어야 할 것이다.