• 인터넷정보학회논문지
• /
• 제24권4호
• /
• pp.127-135
• /
• 2023

본 논문에서는 금융 뉴스 데이터로 추가적인 사전 학습이 진행된 BERT 기반 모델인 FinBERT 모델을 사용하여 금융 영역에서 감성 분석 시 학습시킬 데이터와 그에 맞는 하이퍼파라미터를 찾는 방법을 소개한다. 우리의 목표는 다양한 데이터 세트를 활용하고 하이퍼파라미터를 미세 조정하여 정확한 감성 분석을 위해 FinBERT 모델을 가장 잘 활용하는 방법에 대한 포괄적인 가이드를 제공하는 것이다. 이 연구에서는 제안된 FinBERT 모델 미세 조정 접근법의 아키텍처와 워크플로우를 개괄적으로 설명하고, 감성 분석 태스크를 위한 다양한 데이터 세트와 하이퍼파라미터의 성능을 강조한다. 또한, 감성 라벨링 작업에 GPT-3를 사용함으로써 GPT-3가 적절한 라벨러 역할을 하는지에 대한 신뢰성을 검증한다. 결과적으로 미세 조정된 FinBERT 모델이 다양한 데이터 세트에서 우수한 성능을 발휘 한다는 것을 보여주었고, 각 데이터 세트에 대해 전반적으로 우수한 성능을 보이는 학습률 5e-5와 배치 크기 64의 최적의 조합을 찾았다. 또 일반 도메인의 뉴스보다 일반 도메인의 트위터 데이터 세트에서 성능이 크게 향상됨을 기반으로 금융 뉴스 데이터만으로만 추가적으로 학습시키는 FinBERT 모델에 대한 의구심을 제시한다. 이를 통해 FinBERT 모델에 대한 최적의 접근 방식을 결정하는 복잡한 프로세스를 간소화하고 금융 분야 감성 분석 모델을 위한 추가적인 학습 데이터 세트와 미세 조정 시 하이퍼파라미터 선정에 대한 가이드라인을 제시한다.

• The Journal of Asian Finance, Economics and Business
• /
• 제3권3호
• /
• pp.39-55
• /
• 2016

The rapid expansion of China's trade surplus since the mid-eighties and picking up until the onset of 2008-09 global financial crisis has been a key development in the world economy. While growing trade surplus of China has been viewed with cynicism borne out of an undervalued Yuan and for having being a member of WTO since 2001, many others argue that China's trade surplus reflect changes in China's economic and trade structure and associated shifts in its role within regional and global production chains. We address this issue by analyzing: (i) China's growing and changing trade structure as well as changing structure of trade surplus with the rest of the world, USA, Europe, Japan and rest of Asia, (ii) China's revealed comparative advantage (RCA) with the rest of the world, and (iii) how China's trade policies resulted into a shift in China's trade structure. We find that, not only China has made significant inroads in the world trade since its admission to WTO, but also there has been a noticeable shift in China's trade structure with specialization in high-end technology industries wherein China's exports aided by a well calibrated FDI policy.

• International Journal of Computer Science & Network Security
• /
• 제23권8호
• /
• pp.49-62
• /
• 2023

There has been a rapid increase in the creation and alteration of new malware samples which is a huge financial risk for many organizations. There is a huge demand for improvement in classification and detection mechanisms available today, as some of the old strategies like classification using mac learning algorithms were proved to be useful but cannot perform well in the scalable auto feature extraction scenario. To overcome this there must be a mechanism to automatically analyze malware based on the automatic feature extraction process. For this purpose, the dynamic analysis of real malware executable files has been done to extract useful features like API call sequence and opcode sequence. The use of different hashing techniques has been analyzed to further generate images and convert them into image representable form which will allow us to use more advanced classification approaches to classify huge amounts of images using deep learning approaches. The use of deep learning algorithms like convolutional neural networks enables the classification of malware by converting it into images. These images when fed into the CNN after being converted into the grayscale image will perform comparatively well in case of dynamic changes in malware code as image samples will be changed by few pixels when classified based on a greyscale image. In this work, we used VGG-16 architecture of CNN for experimentation.

• 대한간호학회지
• /
• 제4권1호
• /
• pp.64-80
• /
• 1974

The social understanding of the rehabilitation and welfare work for the handicapped children is nearly devoid in Korea. Especially the medical rehabilitation or nursing for them is left ign. ored as well as the literature preferential to this study is wanted. For the purpose of conducting the study of this thesis Sam Yook Children's Rehabilitation Center, Inc. and Crippled Children's Center, Yonsei University Medical center sampled out; covering three months from Aug. 1 to Oct. 30, 1973, the problems of children's rehabilitation, the characteristics of the children, the medical rehabilitation, nursing, education, and vocational rehabilitation were interviewed and widely grasped. This thesis aimed at developing the rehabilitation, the professional nursing and especially aimed at contributing to the improvement of welfare for the handicapped by analysing, observing the result of this study. The result is as follows: 1. Sam Yook Children's Rehabilitation Center, Inc. has, since its establishment, contributed to the advancement of the crippled children and its purpose consists in implementing services, medical rehabilitation, nursing, education and vocational training. Crippled Children's Center, Yonsei University Medical Center is, being based on Christianity, aiming at the medical treatment and education for the crippled children. 2. All of the children admitted in the children's rehabilitation centers are the crippled children. The ratio of boys io girls is three to two. Boys are more than girls. Orphan formed 55.2 per sent of them. About 60 per cent of them is receiving medical attention free of charge. But there is no orphan in Crippled Children's Center, Yonsei University Medical Center. 3. 15.7 per cent of them have received the previous medical attention before their admitting in the centers; in Sam Yook Children's Rehabilitation Center 8.6 percent, Crippled Children's Center, Yonsei University Medical Center 50 percent; there is remarkable difference between the two. 4. On the standpoint of period of being in the centers, the children who have been over three years in Sam Yook Children's Rehabilitation Center formed 48.7 per cent; in Crippled Children's Center, Yonsei University Medical Center 2.6 percent; there is also considerable difference between the two; they couldn't discharge from the centers owing to the economic conditions and being orphan. 5. Among the diagnosis of the crippled children, poliomyelitis formed highest 51.7 percent of them ; cerebral palsy formed 30 percent secondly in order. Environmental factors (67.8 percent formed about three times of congenital factors (23.7 percent). 6. The children who are capable of doing independently activity of daily living formed 87.9 percent; 73.2 percent can walk about 300 m by wearing brace and so on. The ratio of wearing brace or leg prosthesis formed 47.4 per cent: crutch 44.3 per cent. The medical rehabilitation service and education for the crippled, are comparatively well carried out. But it is desired to improve and cultivate the vocational training, vocational guidance and special nursing to the insufficiency of their implementation. In the tendency that the rehabilitation and welfare for the handicapped are today emphasized, International Handicapped Rehabilitation Association, declaring that 1970s are the decade of rehabilitation, urged to improve positively the policy of rehabilitation and welfare for the handicapped. But here in korea the handicapped, being the object of social prejudice, ignorance, disdain lives in obscurity. Therefore the government or the community should draw up countermeasure on social under-standing, financial support, and rehabilitation services for the handicapped as well as should endeavor to make them participate in social activities as the productive total- person though they are physically imperfect.

• 가족자원경영과 정책
• /
• 제12권2호
• /
• pp.57-72
• /
• 2008

The propose of this study is to examine the economic stresses and coping strategies for elderly households depending on the variables of background, and to analyze the relations between these factors. Elderly households which satisfied the following criteria were recruited for participation: (a) reside in Seoul, or in the metropolitan area (b) live apart from their adult children after retirement and (c) ages over sixty. From September 20th, 2006 to November 30th, 2006, 296 were used for this research. First, the mean score of the economic stress level of elderly households was 2.87 out of 5, and the stress levels of income expenditure and asset debt were intermediate. The mean score of the economic coping strategy was 3.17 out of 5. In order to overcome economic stress, elderly households utilized reducing their expenditure, financial management, and re-employment. The level of elderly households' life satisfaction was 3.29 of 5 and the depression was 3.17 of 5. Second, there was a difference in accordance with the objective economic variables and the degree of the economic stress after the review of the variables of the elderly households and the coping strategy due to economic stress. The result shows that the households which had a low economic status and high economic stress from the objective viewpoint participated in more economic activities. Their reactions were to decrease the overall expenditure through reducing the expenditure rather than to manage the asset effectively through re-employment or to inaugurating a business. Third, I analyzed situational factors, economic stress, and economic coping strategy in order to compare relative contributors to psychological well-being through using regression. At the third phase in the process of analysis, the socio-psychological factors appeared to be significant factors contributing to psychological well-being. Regarding the stress caused by income expenditure increased, when elderly households were more concerned about reducing expenditure and re-employment, their feelings of depression increased.

• 대한건축학회논문집:계획계
• /
• 제34권10호
• /
• pp.131-142
• /
• 2018

Being dealt in Alfred Weber's Theory of the location of Industries, a lot of economic benefits can be obtained through aggregation and clustering of industrial facilities, which derived to the development of industrial complexes in Korea. However, with the IMF economic crisis as well as various institutional changes, the framework of aggregation and clustering of industries is broken, which led to individual developments that took place without any consideration of surrounding industries. For reformation of these condition of industrial complexes, national government-led regeneration projects are being carried out currently. However, national government-led projects mainly focus on profitable projects such as officetel and hotel that are irrelevant to exist composition of industrial complexes which is usually manufacturing base industries and are unable to solve the fundamental problems of industrial complexes. Thus, a necessity of industry clustering is deduced through case analysis of actual private-led manufacturing industry cluster with governance and analysis of benefits on financial, spatial and environmental aspects. In addition, implications on the necessity follow base on factorial analysis on the benefit of clustering development than individual development as well as analysis on the measures taken for successful clustering.

• 한국노년학
• /
• 제38권2호
• /
• pp.271-290
• /
• 2018

이 연구는 사회경제적 박탈 경험이 노인의 자살 생각에 미치는 영향을 분석하는 것을 목적으로 한다. 분석 자료는 한국복지패널 9차년도 자료를 사용하였으며, SPSS를 활용하여 빈도분석, 평균차이분석, 상관분석, 로지스틱 회귀분석 등을 실시하였다. 주요 분석결과는 다음과 같다. 첫째, 공적연금을 받지 못했던 경험, 근로능력이 있으나 실업 상태였던 경험, 경제적 어려움으로 균형 있는 식사를 할 수 없었던 경험, 먹을 것이 떨어졌는데도 더 살 돈이 없었던 경험 등의 순으로 박탈 빈도가 높았다. 둘째, 배우자가 없는 경우, 학력과 소득수준이 낮을수록 자살 생각이 높은 것으로 나타났다. 셋째, 돈이 없어서 본인이나 가족이 병원에 갈수 없었던 경험, 2달 이상 집세가 밀렸거나 집세를 낼 수 없어 집을 옮긴 경험, 먹을 것을 살 돈이 없어 균형 있는 식사를 할 수 없었던 경험, 공과금을 기한 내 납부하지 못한 경험, 근로능력이 있으나 직업을 갖지 못한 경험, 경제적인 어려움 때문에 먹을 것이 떨어졌는데도 더 살 돈이 없었던 경험 등이 노인의 자살 생각에 통계적으로 유의미한 영향을 미치는 것으로 나타났다. 이러한 연구 결과를 토대로 이 논문은 의료급여 사례관리 확대, 노인 주거급여 및 정주 여건 개선, 노인 식생활실태 조사 강화, 노인 공과금 지원 방안 확대 등의 정책 방안을 제시하였다.

• 한국데이타베이스학회:학술대회논문집
• /
• 한국데이타베이스학회 1999년도 춘계공동학술대회: 지식경영과 지식공학
• /
• pp.175-186
• /
• 1999

Detecting the features of significant patterns from their own historical data is so much crucial to good performance specially in time-series forecasting. Recently, a new data filtering method (or multi-scale decomposition) such as wavelet analysis is considered more useful for handling the time-series that contain strong quasi-cyclical components than other methods. The reason is that wavelet analysis theoretically makes much better local information according to different time intervals from the filtered data. Wavelets can process information effectively at different scales. This implies inherent support fer multiresolution analysis, which correlates with time series that exhibit self-similar behavior across different time scales. The specific local properties of wavelets can for example be particularly useful to describe signals with sharp spiky, discontinuous or fractal structure in financial markets based on chaos theory and also allows the removal of noise-dependent high frequencies, while conserving the signal bearing high frequency terms of the signal. To date, the existing studies related to wavelet analysis are increasingly being applied to many different fields. In this study, we focus on several wavelet thresholding criteria or techniques to support multi-signal decomposition methods for financial time series forecasting and apply to forecast Korean Won / U.S. Dollar currency market as a case study. One of the most important problems that has to be solved with the application of the filtering is the correct choice of the filter types and the filter parameters. If the threshold is too small or too large then the wavelet shrinkage estimator will tend to overfit or underfit the data. It is often selected arbitrarily or by adopting a certain theoretical or statistical criteria. Recently, new and versatile techniques have been introduced related to that problem. Our study is to analyze thresholding or filtering methods based on wavelet analysis that use multi-signal decomposition algorithms within the neural network architectures specially in complex financial markets. Secondly, through the comparison with different filtering techniques' results we introduce the present different filtering criteria of wavelet analysis to support the neural network learning optimization and analyze the critical issues related to the optimal filter design problems in wavelet analysis. That is, those issues include finding the optimal filter parameter to extract significant input features for the forecasting model. Finally, from existing theory or experimental viewpoint concerning the criteria of wavelets thresholding parameters we propose the design of the optimal wavelet for representing a given signal useful in forecasting models, specially a well known neural network models.

• 한국지능정보시스템학회:학술대회논문집
• /
• 한국지능정보시스템학회 1999년도 춘계공동학술대회-지식경영과 지식공학
• /
• pp.175-186
• /
• 1999

Detecting the features of significant patterns from their own historical data is so much crucial to good performance specially in time-series forecasting. Recently, a new data filtering method (or multi-scale decomposition) such as wavelet analysis is considered more useful for handling the time-series that contain strong quasi-cyclical components than other methods. The reason is that wavelet analysis theoretically makes much better local information according to different time intervals from the filtered data. Wavelets can process information effectively at different scales. This implies inherent support for multiresolution analysis, which correlates with time series that exhibit self-similar behavior across different time scales. The specific local properties of wavelets can for example be particularly useful to describe signals with sharp spiky, discontinuous or fractal structure in financial markets based on chaos theory and also allows the removal of noise-dependent high frequencies, while conserving the signal bearing high frequency terms of the signal. To data, the existing studies related to wavelet analysis are increasingly being applied to many different fields. In this study, we focus on several wavelet thresholding criteria or techniques to support multi-signal decomposition methods for financial time series forecasting and apply to forecast Korean Won / U.S. Dollar currency market as a case study. One of the most important problems that has to be solved with the application of the filtering is the correct choice of the filter types and the filter parameters. If the threshold is too small or too large then the wavelet shrinkage estimator will tend to overfit or underfit the data. It is often selected arbitrarily or by adopting a certain theoretical or statistical criteria. Recently, new and versatile techniques have been introduced related to that problem. Our study is to analyze thresholding or filtering methods based on wavelet analysis that use multi-signal decomposition algorithms within the neural network architectures specially in complex financial markets. Secondly, through the comparison with different filtering techniques results we introduce the present different filtering criteria of wavelet analysis to support the neural network learning optimization and analyze the critical issues related to the optimal filter design problems in wavelet analysis. That is, those issues include finding the optimal filter parameter to extract significant input features for the forecasting model. Finally, from existing theory or experimental viewpoint concerning the criteria of wavelets thresholding parameters we propose the design of the optimal wavelet for representing a given signal useful in forecasting models, specially a well known neural network models.

• 한국심리학회지 : 문화 및 사회문제
• /
• 제26권3호
• /
• pp.303-327
• /
• 2020

영역 만족도와 삶의 만족도의 상관관계에 대한 연구는 많이 진행되어 왔다. 하지만 다양한 영역 만족도를 종합적으로 고려하여 개별 영역 만족도가 삶의 만족도 및 정서적 안녕감 각각에 미치는 영향력을 살펴본 연구는 부족한 실정이다. 게다가 관련 연구들은 대부분 서구 문화권에서 진행되어 왔으며 문화비교 연구는 많이 이뤄지지 않았다. 이에 본 연구에서는 각 영역 만족도(건강한 라이프스타일, 가족관계, 외모, 재정상황, 학업 성적에 대한 만족도)가 개인의 삶의 만족도 및 정서적 안녕감과 어떤 관계를 가지는지 살펴봄과 동시에, 문화별로 어떤 특정 영역이 삶의 만족도와 정서적 안녕감을 더 잘 예측하는지 한국과 캐나다 두 문화권의 대학생 참가자들을 대상으로 알아보았다. 그 결과, 두 문화에서 다섯 영역의 만족도는 삶의 만족도 및 정서적 안녕감과 정적 상관관계를 보였다. 흥미롭게도, 다섯 영역을 종합적으로 고려했을 때, 문화별로 삶의 만족도와 정서적 안녕감을 예측하는 가장 중요한 요인들에 차이가 있었다. 한국인의 경우 건강한 라이프스타일이 삶의 만족도와 정서적 안녕감을 가장 잘 설명하는 요인으로 나타났다. 반면, 유럽계 캐나다인의 경우 외모 만족도가 삶의 만족도와 정서적 안녕감 모두를 예측하는 가장 중요한 요인으로 나타났으며, 다음으로는 건강한 라이프스타일 만족도가 행복 요소들을 가장 잘 설명하였다. 본 연구 결과는 개인의 주관적 안녕감의 예측에 있어, 문화 공통적으로 영향을 끼치는 삶의 영역이 있으며, 더불어 문화의 특성에 따라 더 좋은 예측력을 가지는 삶의 영역도 있음을 보여준다.