• 제목/요약/키워드: experimental market method

검색결과 132건 처리시간 0.033초

Determining Personal Credit Rating through Voice Analysis: Case of P2P loan borrowers

  • Lee, Sangmin
    • KSII Transactions on Internet and Information Systems (TIIS)
    • /
    • 제15권10호
    • /
    • pp.3627-3641
    • /
    • 2021
  • Fintech, which stands for financial technology, is growing fast globally since the economic crisis hit the United States in 2008. Fintech companies are striving to secure a competitive advantage over existing financial services by providing efficient financial services utilizing the latest technologies. Fintech companies can be classified into several areas according to their business solutions. Among the Fintech sector, peer-to-peer (P2P) lending companies are leading the domestic Fintech industry. P2P lending is a method of lending funds directly to individuals or businesses without an official financial institution participating as an intermediary in the transaction. The rapid growth of P2P lending companies has now reached a level that threatens secondary financial markets. However, as the growth rate increases, so does the potential risk factor. In addition to government laws to protect and regulate P2P lending, further measures to reduce the risk of P2P lending accidents have yet to keep up with the pace of market growth. Since most P2P lenders do not implement their own credit rating system, they rely on personal credit scores provided by credit rating agencies such as the NICE credit information service in Korea. However, it is hard for P2P lending companies to figure out the intentional loan default of the borrower since most borrowers' credit scores are not excellent. This study analyzed the voices of telephone conversation between the loan consultant and the borrower in order to verify if it is applicable to determine the personal credit score. Experimental results show that the change in pitch frequency and change in voice pitch frequency can be reliably identified, and this difference can be used to predict the loan defaults or use it to determine the underlying default risk. It has also been shown that parameters extracted from sample voice data can be used as a determinant for classifying the level of personal credit ratings.

유해물질 규제법(RoHS)에 따른 휴대폰 내의 중금속 함유량 측정을 위한 스크리닝법 연구 (Study on the screening method for determination of heavy metals in cellular phone for the restrictions on the use of certain hazardous substances (RoHS))

  • 김영혜;이재석;임흥빈
    • 분석과학
    • /
    • 제23권1호
    • /
    • pp.1-14
    • /
    • 2010
  • 모든 전기전자제품에 대한 유해 중금속 규제법(RoHS)이 유럽과 중국 등 세계 각 나라에서 시작 되면서 이에 대한 중요성이 대두되고 있다. 이와 관련하여, 현재 세계 전기전자 협회(IEC)에서 발표된 IEC 62321 문건은 기존의 표준 분석 규격들과 마찰이 있을 수 있다. 반면에, IEC기술위원회(TC 111) 에서 발표된 시료 채취 및 처리 방법(sampling)에 관한 일반공개규격(Publicly Accessible Specification: PAS)은 기존 규격들과 상호보완적으로 응용될 수 있다. 이 실험에서는 PAS에 준하여 휴대폰시료를 분리 및 분해하는 방법을 찾고, 기존의 분석장치를 이용한 스크리닝방법과 비교하고자 하였다. X선 형광분석법(XRF)은 시료의 전 처리가 필요 없어서, 신속한 분석이 가능하므로 스크리닝에 탁월한 기능을 보여준다. 하지만 이 방법은 표면의 유해물질 정보만 알 수 있으며, 심각한 매트릭스 간섭과 정량을 위한 다양한 표준물질이 없다는 제약이 따르기 때문에, 이를 보완할 수 있는 방법이 필요하다. 본 연구에서는 레이저 박리 유도결합플라스마 질량분석장치(LA-ICP-MS), 에너지 분산형 XRF (ED-XRF) 와 전자주사현미경 에너지분산 X선분광기(SEM-EDX)를 휴대폰 스크리닝에 적용하여 보았으며, 유해중금속이 검출된 일부 부품의 경우에는 흑연로 원자흡수분광분석법(GF-AAS)을 수행하여 농도를 측정하였다. 이 실험결과, 배터리 일부 부품의 경우, GF-AAS와 XRF의 Pb 측정결과는 각각 0.92%와 5.67%로서 차이가 많이 나는 것을 알 수 있었다. 또한, XRF의 상대편차 범위 23-168%는 LA-ICP-MS의 편차 범위인 1.9-92.3% 보다 월등히 큰 것으로 나타났다. 결론적으로, 스크리닝 목적으로 XRF 분석방법을 이용하였다 할지라도, 정확한 함량을 얻기 위하여 GF-AAS의 분석을 수행해야 할 것으로 판단된다. 그리고, 기존 IEC 문건의 스크리닝 방법에는 언급되어 있지 않으나, 본 연구에서는 LA-ICP-MS가 전기전자제품 내에 있는 유해물질들에 대한 정보를 신속하게 얻을 수 있는 스크리닝 방법으로 활용될 수 있음을 확인하였다.

백미대맥(白米大麥) 및 속(粟) 사료(飼料)가 실험용(實驗用) 백서(白鼠)에 미치는 영양학적(營養學的) 및 생물화학적(生物化學的) 연구(硏究) (Biochemical and Nutritional Effects of Rice, Barley, and Millet on Rats)

  • 허검
    • Journal of Nutrition and Health
    • /
    • 제1권1호
    • /
    • pp.9-18
    • /
    • 1968
  • Dividing into six(6) experimental groups; Stock Diet(SD) Group, Rice Diet(RD) Group, Rice Diet supplemented with Barley(RD-BD) Group, Barley Diet(BD) Group, Rice Diet supplemented with Millet(RD-MD) Group, and Millet Diet(MD) Group, biochemical and nutritional effects of rice, barley, and millet on rats are studied. The grains used in this experiment are polished ones and purchased from market. Six(6) weaning male albino rats are housed into the individual cages in each group and fed with each experimental diet and water ad libitum for 21 weeks. The composition of the experimental diets are shown in Table 1. At the end of experiment, the appearences of rats are observed and the bloods are sampled by heart puncture method following ether anesthesia for determinations of serum protein and serum cholesterol. And the livers are sampled immediately right after heart after heart puncture for determinations of liver weight, liver nitrogen, liver fat, and liver glycogen. And also liver tissues are observed histologically. The result of growth experiment is shown in Table 2 and Figure 7. The results of blood analysis are shown in Table 3 and the results of liver analysis are shown in Table 4. And the histological observations of liver tissue are shown in Figures 1, 2, 3, 4, 5, and 6. According to the above results, it could be summarized as follows; 1) The normal appearences are observed at SD Group and BD Group. Alopecia is observed at RD, RD-­BD, RD-MD, and MD groups and the wet appearences of hair are observed at RD-MD and MD groups. The color of hair is changed to yellowish white at RD, RD-MD groups. The maximum growth inhibition, is observed at MD group and the growths of RD, RD-BD, BD, and RD-MD groups are significantly lower than that of SD group. The growth of RD group is lower than those of RD-BD and BD groups, though this difference is not significant. 2) At the blood analysis, the highest serum protein content is shown at SD group and the lowest serum protein is observed at MD group. The serum cholesterol contents are not significantly different among the six(6) experimental groups. 3) The highest liver weight is observed at SD group and the lowest weight is shown at MD group. However, oppositely the highest liver weight per 100 grams of body weight at MD group and the lowest liver weight per 100 grams of body weight at SD group are observed as shown in Table 6. 4) The differences of liver nitrogen contents of RD, RB-BD, BD, RD-MD, and MD Groups are not significant except SD group showing the highest content. 5) The significant highest content of liver fat is observed at RD group and the lowest content of liver fat is shown at SD group. And no significant differences in contents of liver glycogen are observed among the six(6) experimental groups. 6) No significant abnormal histological changes of liver tissue are observed at SD, RD-BD, and BD groups. However, the fat metamorphosis at peripheral zone of liver tissue is observed at RD Group and the some histological abnormal changes are observed at RD-MD and MD groups. According to the above summaries, the followings might be estimated and recommended; 1) The BD group shows histological normality of liver tissue, improvement of liver fat and liver nitrogen, normality of appearence, and better growth compared with RD Group. This could be indicated that barley is nutritionally and biochemically better than rice as the feeding diet on rat. And it might be considered that there might be the species speciality for nutritionary evaluation of food. 2) Recently, Marshall indicated that the high consumption of carbohydrate causes fatty liver and Kondo reported that the higher consumption of rice causes the more heart disease and the shorter life span. Considering these reports, the higher consumption of rice as a main dish in our country should be considered as an important national nutritional problem. And this might be improved by supplementing with barley or other grains to rice.

  • PDF

유중건조를 이용한 축산분뇨슬러지의 고형연료화 공정 연구 (A study on the RDF(Refuse Derived Fuel) making process of Livestock manure sludge by oil-drying method)

  • 이준호;박소연;이경호;하진욱
    • 한국산학기술학회논문지
    • /
    • 제18권2호
    • /
    • pp.294-301
    • /
    • 2017
  • 본 연구에서는 유중건조 공정으로 가축분뇨 슬러지 고형 연료를 제조하기 위한 최적 조건을 선정하는 연구를 수행하였다. 연구결과물(슬러지 고형 연료)의 특성 평가를 위하여 유증 증발량 측정 실험, 슬러지 유중건조 실험, 그리고 건조 슬러지 성형 실험을 수행하였으며, 칼로리미터와 PXRF 장비를 이용하여 고형 연료의 성능을 측정하였다. 또한, 현재 시중에서 일반적으로 사용되고 있는 목재 고형 연료와 본 연구에서 개발한 슬러지 고형 연료 사이의 열량 비교 연구를 수행하였다. 실험결과에 따르면 대량생산 및 환경 친화적 측면을 함께 고려하였을 때, 30g의 슬러지를 $130^{\circ}C$의 식물성 기름으로 25분간 처리하는 것이 고형연료 생산에 가장 적합한 조건으로 나타났다. 본 연구에서 제조한 슬러지 고형 연료의 열량은 5211kcal/kg이었으며, 이는 시중에서 널리 사용되고 있는 목재 고형 연료의 열량보다 큰 값이다. 마지막으로, PXRF 측정 결과 황을 제외한 중금속은 검출되지 않았다. 그러므로 추후 산업용 제조기술로의 개발 시 황 성분 제어 공정에 대한 추가적인 연구가 필요할 것으로 보인다.

한국수산업의 해외진출전략에 관한 연구 (Foreign Entry Strategies for Korean Fishery Firms)

  • 김회천
    • 수산경영론집
    • /
    • 제15권1호
    • /
    • pp.131-153
    • /
    • 1984
  • Fishery resources are still abundant compared with other resources and the possibility of exploitation is probably great. The Korean fishery industry has grown remarkably since 1957, and Korea is ranked as one of the major fishery countries. Its of fishery products reached the 9th in the world and the value of exports was 5th in 1982. But recently a growth rate has slowed down, due to the enlargement of territorial seas by the declaration of the 200 mile, Exclusive Economic Zone, the tendency to develop fishery resources strate-gically in international bargaining, the change in function of the international organizations, the expansion of regulated waters, the illegal arrest of our fishing boats, the rapid rise in oil prices, and the fall in fish prices, the development of fishery resources as a symbol of nationalism, the fishing boats decreptitude, the rise of crew wages, regulations on fishing methods, fish species, fishing season, size of fish, and mesh size, fishing quotas and the demand of excessive fishing royalties. Besides the the obligation of coastal countries, employing crews of their host countries is also an example of the change in the international environment which causes the aggravation of foreign profit of fishing firms. To ameliorate the situation, our Korean fishery firms must prepare efficient plans and study systematically to internationalize themselves because such existing methods as conventional fishing entry and licence fishing entry are likely to be unable to cope with international environmental change. Thus, after the systematic analysis of the problem, some new combined alternatives might be proposed. These are some of the new schemes to support this plan showing the orientation of our national policy: 1. Most of the coastal states, to cope with rapid international environmental change and to survive in the new era of ocean order, have rationalized their higher governmental structure concerning the fishery industries. And the coastal countries which are the objectives of our expecting entry, demand excessive economic and technical aid, limit the number of fishing boats’entry and the use of our foreign fishing bases, and regulate the membership of the international fishery commissions. Especially, most of the coastal or island countries are recently independent states, which are poorer in national budget, depend largely on fishing royalties and licence entry fees as their main resources of national finance. 2. Alternatives to our entry to deep sea fishing, as internationalization strategies, are by direct foreign investment method. About 30 firms have already invested approximately US $ 8 million in 9 coastal countries. Areas of investment comprise the southern part of the Atlantic Ocean, the Moroccan sea and five other sea areas. Trawling, tuna purse seining and five other fields are covered by the investment. Joint-venture is the most prominent method of this direct investment. If we consider the number of entry firms, the host countries, the number of seas available and the size of investment, this method of cooperation is perhaps insufficient so far. Our fishery firms suffer from a weakness in international competitive ability, an insufficiency of information, of short funds, incompetency in the market, the unfriendliness of host coastal countries, the incapability of partners in joint-ventures and the political instability of the host countries. To enlarge our foreign fishing grounds, we are to actively adopt the direct investment entry method and to diversity our collaboraboration with partner countries. Consequently, besides proper fishing, we might utilize forward integration strategies, including the processing fied. a. The enterprise emigration method is likely to be successful in Argentina. It includes the development of Argentinian fishing grounds which are still not exploited in spite of abundant resources. Besides, Arentina could also be developed as a base for the exploitation of the krill resources and for further entries into collaboration with other Latin American countries. b. The co-business contract fishing method works in American territorial seas where American fishermen sell their fishery products to our factory ships at sea. This method contributes greatly to obtaining more fishing quotas and in innovation bottom fishing operation. Therefore we may apply this method to other countres to diffuse our foreign fishing entry. c. The new fishing ground development method was begun in 1957 by tuna long-line experimental fishing in the Indian Ocean. It has five fields, trawling, skipjack pole fishing and shrimp trawling, and so on. Recently, Korean fisheries were successful in the development of the Antarctic Ocean krill and tuna purse seining. 3. The acceleration of the internationalization of deep sea fishing; a. Intense information exchange activities and commission participation are likely to be continues as our contributions to the international fishery organizations. We should try to enter international fishery commissions in which we are not so far participating. And we have to reform adequately to meet the changes of the function of the international commissions. With our partner countries, we ought to conclude bilateral fishery agreements, thus enlarging our collaboration. b. Our government should offer economic and technical aids to host countries to facilitate our firms’fishery entry and activities. c. To accelerate technical innovation, our fishery firms must invest greater amount in technical innovation, at the same time be more discriminatory in importing exogeneous fishery technologies. As for fishing methods; expanded use of multi-purpose fishing boats and introduction of automation should be encuraged to prevent seasonal fluctuations in fishery outputs. d. The government should increases financial and tax aid to Korean firms in order to elevate already weak financial structure of Korean fishery firms. e. Finally, the government ought to revise foreign exchange regulations being applied to deep sea fishery firms. Furthermore, dutes levied on foreign purchaed equipments and supplies used by our deep sea fishing boats thould be reduced or exempted. when the fish caught by Korean partner of joint-venture firms is sold at the home port, pusan, import duty should be exempted.

  • PDF

집중형센터를 가진 역물류네트워크 평가 : 혼합형 유전알고리즘 접근법 (Evaluating Reverse Logistics Networks with Centralized Centers : Hybrid Genetic Algorithm Approach)

  • 윤영수
    • 지능정보연구
    • /
    • 제19권4호
    • /
    • pp.55-79
    • /
    • 2013
  • 본 연구에서는 집중형 센터를 가진 역물류네트워크(Reverse logistics network with centralized centers : RLNCC)를 효율적을 해결하기 위한 혼합형 유전알고리즘(Hybrid genetic algorithm : HGA) 접근법을 제안한다. 제안된 HGA에서는 유전알고리즘(Genetic algorithm : GA)이 주요한 알고리즘으로 사용되며, GA 실행을 위해 0 혹은 1의 값을 가질 수 있는 새로운 비트스트링 표현구조(Bit-string representation scheme), Gen and Chang(1997)이 제안한 확장샘플링공간에서의 우수해 선택전략(Elitist strategy in enlarged sampling space) 2점 교차변이 연산자(Two-point crossover operator), 랜덤 돌연변이 연산자(Random mutation operator)가 사용된다. 또한 HGA에서는 혼합형 개념 적용을 위해 Michalewicz(1994)가 제안한 반복적언덕오르기법(Iterative hill climbing method : IHCM)이 사용된다. IHCM은 지역적 탐색기법(Local search technique) 중의 하나로서 GA탐색과정에 의해 수렴된 탐색공간에 대해 정밀하게 탐색을 실시한다. RLNCC는 역물류 네트워크에서 수집센터(Collection center), 재제조센터(Remanufacturing center), 재분배센터(Redistribution center), 2차 시장(Secondary market)으로 구성되며, 이들 각 센터 및 2차 시장들 중에서 하나의 센터 및 2차 시장만 개설되는 형태를 가지고 있다. 이러한 형태의 RLNCC는 혼합정수계획법(Mixed integer programming : MIP)모델로 표현되며, MIP 모델은 수송비용, 고정비용, 제품처리비용의 총합을 최소화하는 목적함수를 가지고 있다. 수송비용은 각 센터와 2차 시장 간에 제품수송에서 발생하는 비용을 의미하며, 고정비용은 각 센터 및 2차 시장의 개설여부에 따라 결정된다. 예를 들어 만일 세 개의 수집센터(수집센터 1, 2, 3의 개설비용이 각각 10.5, 12.1, 8.9)가 고려되고, 이 중에서 수집센터 1이 개설되고, 나머지 수집센터 2, 3은 개설되지 않을 경우, 전체고정비용은 10.5가 된다. 제품처리비용은 고객으로부터 회수된 제품을 각 센터 및 2차 시장에서 처리할 경우에 발생되는 비용을 의미한다. 수치실험에서는 본 연구에서 제안된 HGA접근법과 Yun(2013)의 연구에서 제안한 GA접근법이 다양한 수행도 평가 척도에 의해 서로 비교, 분석된다. Yun(2013)이 제안한 GA는 HGA에서 사용되는 IHCM과 같은 지역적탐색기법을 가지지 않는 접근법이다. 이들 두 접근법에서 동일한 조건의 실험을 위해 총세대수 : 10,000, 집단의 크기 : 20, 교차변이 확률 : 0.5, 돌연변이 확률 : 0.1, IHCM을 위한 탐색범위 : 2.0이 사용되며, 탐색의 랜덤성을 제거하기 위해 총 20번의 반복실행이 이루어 졌다. 사례로 제시된 두 가지 형태의 RLNCC에 대해 GA와 HGA가 각각 실행되었으며, 그 실험결과는 본 연구에서 제안된 HGA가 기존의 접근법인 GA보다 더 우수하다는 것이 증명되었다. 다만 본 연구에서는 비교적 규모가 작은 RLNCC만을 고려하였기에 추후 연구에서는 보다 규모가 큰 RLNCC에 대해 비교분석이 이루어 져야 할 것이다.

미국산 쇠고기 수입재개 발표에 따른 소비자 후생손실 측정 (Measuring Consumers' Welfare Losses due to Announcement of Resuming US-Beef Imports)

  • 엄영숙
    • 자원ㆍ환경경제연구
    • /
    • 제18권3호
    • /
    • pp.495-521
    • /
    • 2009
  • 본고는 식품위험이 현실화되지 않은 상황에서 건강위해에 대한 심리적 불안과 우려로 인하여 발생할 수 있는 소비감소와 후생손실을 2008년 4월 미국산 쇠고기 수입재개 발표에 따른 BSE 위험을 사례로 분석하였다. 조건부가치측정법이나 실험시장접근법을 사용하여 식품안전에 대한 소비자들의 단위당 지불의사나 가격프리미엄을 도출했던 기존의 연구들과는 달리, 소비자들이 기록한 쇠고기 구입량, BSE 위험에 대한 주관적 심각성, 표본내 개인들이 직면한 가격지수 등을 산정하여 개인별 쇠고기 수요함수를 추정하고 소비자들이 주관적으로 체험하는 정확한 후생효과인 동등변화(EV)와 소비자잉여의 변화를 측정함으로써 소비자들이 식품안전성에 부여하는 가치를 측정하였다. 2008년 10월 말부터 11월 중순까지 쇠고기를 소비하는 360명의 주부들을 대상으로 실시한 설문조사 자료를 사용하여 쇠고기 수요함수를 추정한 결과, 수입재개 발표 전 후의 경우 모두 자체 가격, 대체재의 가격, 그리고 소득변수 등의 부호가 경제이론에 부합되었고 통계적으로 유의하였으며 발표 전 후 두 기간 동안에 큰 차이가 관찰되지 않았다. 반면에 수업재개 발표 이후 형성된 BSE 위험에 대한 심각성을 나타내는 변수는 발표 이후의 수요함수에서는 통계적으로 유의한 부의 부호를 가진 반면에, 발표 이전의 수요함수에서는 통계적으로 유의하지 않았다. 그리하여 수입재개 발표 후의 BSE 위험에 대한 우려증가로 인한 품질저하 인식에 따른 소비자들의 후생손실을 측정한 결과 가구당 30,000원 정도로 계산되었다. 또한 이 정도의 후생손실을 가져올 BSE 위험에 대한 우려수준은 수입재개 전에 비하여 35% 이상 증가한 것으로 나타났다.

  • PDF

Wavelet Thresholding Techniques to Support Multi-Scale Decomposition for Financial Forecasting Systems

  • Shin, Taeksoo;Han, Ingoo
    • 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.

  • PDF

Wavelet Thresholding Techniques to Support Multi-Scale Decomposition for Financial Forecasting Systems

  • Shin, Taek-Soo;Han, In-Goo
    • 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.

  • PDF

범주형 속성 기반 군집화를 위한 새로운 유사 측도 (A New Similarity Measure for Categorical Attribute-Based Clustering)

  • 김민;전주혁;우경구;김명호
    • 한국정보과학회논문지:데이타베이스
    • /
    • 제37권2호
    • /
    • pp.71-81
    • /
    • 2010
  • 데이터의 군집을 찾아내는 문제는 패턴 인식, 이미지 처리, 시장 조사 등 많은 응용 분야에서 널리 사용되고 있다. 군집의 질을 결정하는 핵심 요소로는 유사 측도, 차원의 개수 등이 있다. 유사 측도는 데이터의 특성을 반영하여 다르게 정의되어야 하는데, 대부분 기존의 연구들은 데이터를 특징 지어주는 속성이 수치형으로 주어진 경우에 국한되어 있었다. 속성이 범주형으로 주어진 경우도 실생활에 많이 존재하지만, 범주형 변수에 대한 속성값의 유사성은 값의 순서가 고유하게 정해지지 않아서 정의하기 어렵다. 이에 더하여, 고차원 데이터에 대해서는 데이터 점들이 희박하게 위치하여 가까운 점과 먼 점간의 차이가 거의 없고, 군집화 결과가 좋지 않을 수 있다. 이 문제를 해결하기 위해 부분 차원 군집화 방법이 제안되어 왔다. 부분 차원 군집화 방법은 각 군집을 발견하기에 적합한 부분 차원을 선택하면서 군집화를 수행하는 방법이다. 본 논문에서는 범주형 속성으로 특징지어진 고차원 데이터를 부분 차원 군집화하기 위한 새로운 유사 측도를 제안한다. 유사 측도는 각 군집은 다른 군집과 구별되는 특정 정보를 잘 표현할 수 있어야 한다는 기본적인 가정 하에 속성들 사이의 상관성을 반영하여 정의되었다. 이들 모두를 반영한 유사측도는 기존에 존재하지 않았다는 점에서 본 연구는 의미가 있다. 실제 데이터 집합을 군집화하는 실험을 통해 제안하는 방법이 다른 군집화 방법보다 저차원 데이터와 고차원 데이터 모두에 대해 좀 더 정확한 군집 결과를 얻을 수 있음을 보였다.