• Water for future
• /
• v.26 no.1
• /
• pp.51-62
• /
• 1993

The most significant factor in estimating annual runoff must be the precipitation. But in the previous study, the watershed area instead of precitation was included as an independent variable in regression model in the process of checking accurate data. The criterion of accurate data was the runoff ratio in the range of 20% to 100%. In this study the valid range of evapotranspiration was adopted as a criterion of accurate data and the same data were reexamined. It came up with following model which has a high coefficient of determination and conforms to hydrologic theory. R=-518.25+0.8834P where, R: runoff depth(mm) P: precipitation(mm) This regression model was found to be stable by cross-validation and is proposed as annual runoff estimation model applicable to ungaged small and medium watersheds in Korea.

• 한국벤처창업학회:학술대회논문집
• /
• 2023.04a
• /
• pp.145-148
• /
• 2023

최근 딥테크 및 글로벌 스타트업들의 기술경쟁력 제고를 위한 국가적 지원이 증대함에 따라 원천기술을 보유한 기술창업 건 수역대급으로 증가하고 있는 추세이다. 기술 기업들은 사업의 다각화 및 글로벌 진출에 유리하다는 장점이 있으나, 이들의 전략 성공을 위해서는 체계적인 인큐베이팅 및 자금유치는 필수적이라 할 수 있다. 본 연구는 첨단 가술스타트업의 투자유치성과에 중요하게 여겨지는 요인들이 IT 및 BT 산업별로 어떻게 다르게 영향을 미치고 있는지를 연구하는 목적으로 하고 있다. 선행연구에 따르면 기술 스타트업에 대해서는 그들이 보유한 기술 역량(특허), 창업자/팀의 능력, 제휴 등을 중요한 자원으로 보고 있으며, 이는 기업의 퀄리티를 차별화하는 신호로 작용하여 투자유치에 긍정적 영향을 미치는 것으로 나타났다. 본 연구에서는 국내 기술 스타트업의 산업별 차이에 따라 기술 스타트업의 특성이 투자유치성가에 미치는 영향을 비교분석하고자 한다. 본 연구는 2022년 기준 서울시 창업지원시설에 입주한 스타트업 122개사를 대상으로, 전 산업을 대상으로 한 회귀분석모델과 글로벌 산업기술 분류 기준에 따라 구분된 BT 및 IT 산업만을 대상으로 한 회귀분석 모델을 설정하여 연구목적을 규명하고자 한다. 본 연구 결과는 국내 기술스타트업의 투자유치 의사결정에 대한 실증분석을 통하여 기존 이론을 검증하였다는 것에 의의가 있으며, 산업별로 투자유치에서 중요하게 고려하는 요소들이 다르게 작용할 수 있음을 밝히고 산업별 특성에 따라 투자유치 시 실질적으로 필요한 요소들을 제시한다는데 실무적인 시사점을 지니고 있다.

• Journal of the Korea Concrete Institute
• /
• v.27 no.6
• /
• pp.669-676
• /
• 2015

Time dependent model for prediction of compressive strength development of concrete cured by microwave heating form was presented in this study. The presented model is similar to the equation which is given in ACI 209R-92 but the constants which is dependent on cement type and curing method in the presented model are modified by the regression analysis of the experimental data. Laboratory scale concrete specimens were cast and cured by the microwave heating form and drilled cores extracted from the specimens were fractured in compression. The measured core strengths are converted to standard core and in-situ strengths. These in-situ strengths are used for the regression.

• KIPS Transactions on Computer and Communication Systems
• /
• v.5 no.10
• /
• pp.293-302
• /
• 2016

Since the electricity is produced and consumed simultaneously, predicting the electric load and securing affordable electric power are necessary for reliable electric power supply. In particular, a university campus is one of the highest power consuming institutions and tends to have a wide variation of electric load depending on time and environment. For these reasons, an accurate electric load forecasting method that can predict power consumption in real-time is required for efficient power supply and management. Even though various influencing factors of power consumption have been discovered for the educational institutions by analyzing power consumption patterns and usage cases, further studies are required for the quantitative prediction of electric load. In this paper, we build an electric load forecasting model by implementing and evaluating various machine learning algorithms. To do that, we consider three building clusters in a campus and collect their power consumption every 15 minutes for more than one year. In the preprocessing, features are represented by considering periodic characteristic of the data and principal component analysis is performed for the features. In order to train the electric load forecasting model, we employ both artificial neural network and support vector machine. We evaluate the prediction performance of each forecasting model by 5-fold cross-validation and compare the prediction result to real electric load.

• Journal of dental hygiene science
• /
• v.9 no.4
• /
• pp.419-426
• /
• 2009

The occurrence of toothache signals the malfunction in oral health, which allows the detection of any abnormal condition in the oral cavity at an early stage to prevent the condition from worsening, and thus can act as a preventive measure. This study has looked into the status of oral health management in relation to toothache through the structured survey administered to 235 college students. Based on the survey results, this study aimed at comparing the toothache occurrence prediction between regression analysis and CART model in order to clarify the relationship between the factors of oral health management habits that contribute to toothache occurrence. According to the result, there was a difference between the present health status and the health status of the past year depending on the presence or non-presence of toothache occurrence (p<0.05). There was a difference in the regularity of meal time depending on the presence non-presence of toothache occurrence from the dietary habits of the research subjects (p<0.05). As for the presence or non-presence of toothache occurrence from the oral hygiene habits of the research subject, there was a difference between the occurrence and nonoccurrence of bleeding during brushing or flossing (p<0.05). According to the results of regression analysis, no factors were signifiant in the relationship with the presence or non-presence of toothache occurrence from the status of life habits and oral hygiene habits. 70% of the researched group was randomly selected as the sample for generating an analytical model and the remaining 30% was used as the sample for generating an evaluation model. According to the results of CART model, the occurrence of toothache was higher in the case of irregular meal time and poor current health condition than the case of average or satisfactory health condition. The above results imply that CART model is very useful technique in predicting toothache occurrence compared to regression analysis, and suggests that CART model could be very useful in predicting other oral diseases including toothache.

• Journal of Dental Rehabilitation and Applied Science
• /
• v.24 no.3
• /
• pp.299-310
• /
• 2008

In this study, the regression analysis method was tested for the estimation of peak stress at stress concentration area in the cervical bone. Submerge type EZ plus implant (Megagen. Daegu, Korea), 4.1 mm in cervical diameter and 9.6 mm in endosseous length, were axisymmetrically modelled together with surrounding alveolar bone of which the width was 10 mm. Vertical force of 100 N was applied to a head of crown above 8.5 mm from the outer surface of the cortical bone. Four different mesh models were composed of differently sized elements in vicinity of sharp corners, and they include 6 stress monitoring points that are located in the same geometrical points regardless of the differences in the meshes. Primary consideration was given to the stresses in the cortical bone surrounding the implant neck. The results showed that virtually all the stresses were concentrated in the cortical bone regardless of mesh designs. The peak stresses were successfully calculated by a regression analysis in a stable manner, as far as the mesh is designed to represent the acute gradient of stresses near the sharp corner.

• 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.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2017.05a
• /
• pp.6-6
• /
• 2017

본 연구의 목적은 독립적으로 수행되어진 하천 복원 가치추정 선행연구들을 대상으로 메타회귀분석을 활용한 편익이전을 제시함으로써, 편익이전의 타당성 및 적용 가능성을 검토하는 데 있다. 문헌검색을 통해 '하천 가치평가', '하천 가치', '물 환경 가치추정', '하천 편익' 등에 관한 자료를 수집하였으며, 총 60편의 연구를 수집했다. 그 중 중복된 연구를 제외하고 가치추정 단위를 '원/년/가구'로 통일함으로써 51편의 연구를 분석에 사용했으며, 90개의 가치 추정치가 실증분석에 사용되었다. 본 연구는 국내에서 수행된 하천 복원 가치 추정연구를 집대성하여 DB를 구축하고 요약통계량을 중심으로 선행연구 결과를 기술하였으며, 메타회귀분석을 실시한 후, policy site의 특성과 조건에 맞게 함수를 조정하고, 조정된 함수를 사용하여 policy site의 가치를 예측하였다. 종속변수로는 총 가치(원/년/가구, 2015년 불변가격)가, 독립변수로는 하천유형, 위치, 규모, 환경 서비스특성, 그리고 방법론 특성, 지불형태, 대상지 사회경제적 특성 변수들이 포함되었다. 모형의 추정결과 조정된 값은.420으로써 종속변수 총변이의 42.0%를 모형이 설명하는 것으로 나타났다. 메타회귀분석을 통해 본류에서 멀어지는 소규모하천일수록 하천의 경제적 가치를 더 크게 느끼는 것으로 나타났으며, 전체적인 영향력 크기를 고려해 본다면 하천을 복원할 때 수질정화기능, 서식지기능, 이수기능, 치수기능, 여가 및 수변공간으로의 기능 순으로 고려하는 것이 하천의 가치를 보다 높일 수 있을 것으로 보였다. 또한 지불방법은 매월, 인당 지불하는 것으로 제시할 때 경제적 가치 추정치를 높일 수 있는 것으로 해석되었다. 모델추정 결과를 활용한 함수이전에서는 만경강의 특성을 반영하고 조정함으로써 만경강의 가치를 추정하였으며, 모형으로부터 얻은 만경강 가치 예측치는 가구당 매년 41,214원으로 추정되었다. 본 연구의 메타회귀분석은 선행연구를 객관적으로 종합할 수 있는 분석의 틀로서 충분한 활용 타당성이 인정되는 것으로 보이며, 편익이전 시에 policy site의 자원특성과 조건에 맞춰 함수를 조정하여 예측치를 제시함으로써 메타회귀분석 함수이전의 융통성을 보여주었다. 이에 메타회귀분석을 통한 편익이전은 타당성 및 적용 가능성 측면에서 긍정적으로 판단된다.

• Journal of the Korea Concrete Institute
• /
• v.22 no.4
• /
• pp.481-488
• /
• 2010

Since existing constitutive models developed for confined normal strength concrete overestimate ductility when they are applied to confined high strength concrete, these models cannot be directly applied to confined high strength concrete. In an effort to solve this problem, an accurate stress-strain relationship of the hihg strength concrete needs to be formulated by examining the confinement effects due to increase of the concrete strength. In this study, a constitutive model is developed to express the stress-strain relationship of confined high strength concrete by carrying out regression analysis of the main parameters affection strength and ductile behavior of reinforced high strength concrete columns. Twenty-five test specimens were chosen from the reported experimental studies in the literature. The experimental results of stress-strain relationships of show a good agreement with results of the stress-strain relationships of suggested high strength concrete, covering a strength range between 60 and 124 MPa.

• Proceedings of the KIEE Conference
• /
• 2002.07d
• /
• pp.2749-2751
• /
• 2002

일반화된 회귀 신경망 (GRNN)을 이용하여 플라즈마을 분류하는 새로운 알고리즘을 보고한다. 데이터분포를 통계적인 평균치와 표준편차를 이용하여 특징지었으며, 바이어스 인자을 이용하여 9 종류의 데이터을 발생하였다. 각 데이터에 대하여 GRNN의 학습인자를 최적화하였으며, 모델성능은 예측과 분류 정확도로 나누어 바이어스와 학습인자의 함수로 분석하였다. 바이어스는 모델성능에 상당한 영향을 주었으며, 학습인자와의 상호작용을 통하여 완전 분류를 이루었다.