• 제목/요약/키워드: over-fitting

검색결과 347건 처리시간 0.025초

여자 청소년용 가상모델 개발을 위한 체형구분 및 설계방법 연구 (An Analysis of Young Girls' Somatotype and the Design for Virtual Fitting Model)

  • 강여선
    • 한국의류학회지
    • /
    • 제41권6호
    • /
    • pp.1109-1123
    • /
    • 2017
  • This study analyzed a somatotype of teenager's that was suitable to improve the reality of a virtual model size. We analyzed 843 teenagers 12-18 years old from the 6th Size Korea data. First, factor analysis was done for abstracting new criteria and dividing the somatotype; subsequently, we selected the waist height proportion to stature (body proportion) and drop (torso shape). Next, the cluster analysis was done with these criteria; subsequently, 5 body proportion types and 7 torso shapes were distinguished. A virtual model size for 4 somatotype with more than 50 persons was also designed by a regression analysis that constituted sizes for each factor. The designed model size was compared with body size as well as with Clo's virtual model size. The research model showed a high similarity in sizes with body as well as improved reality over the Clo model that presented size problems such as low waist height, bigger bust, and smaller thigh circumference than the real body.

X-Ray Reflectivity Analysis Incorporated with Genetic Algorithm to Analyze the Y- to X Type Transition in CdA LB Film

  • 최정우;조경상;이희우;이원홍;이한섭
    • Bulletin of the Korean Chemical Society
    • /
    • 제19권5호
    • /
    • pp.549-553
    • /
    • 1998
  • The structure and layer distribution of cadmium arachidate Langmuir-Blodgett film were analyzed by the small angle X-ray reflectivity measurements using synchrotron radiation. Y-to X type transition was ocurred during the 39th passage of deposition of cadmium arachidate. Based on the measurement of the consumed area of the monolayer, it was determined that about 27.5 layer was deposited. Using the synchrotron X-ray, the reflectivity profile of cadmium arachidate LB film over the wide range of grazing angle was obtained. The X-ray reflectivity profile was analyzed using the recursion formula. By fitting the location and dispersion of the subsidiary maxima between the Bragg peaks of the measured reflectivity profile with that of the calculated reflectivity profile, the average thickness and the distribution of layer thickness were evaluated. The genetic algorithm was adopted to the fitting of reflectivity profile to evaluate the optimum value of the number distribution of layer. Based on the morphology measurement with an atomic force microscopy (AFM), the domain structure and mean roughness of LB films were obtained. The mean roughness value calculated based on the number of layer distribution obtained from the measurement by AFM is consistent with that obtained from X-ray reflectivity analysis.

Volume Transport on the Texas-Louisiana Continental Shelf

  • Cho Kwang-Woo
    • Fisheries and Aquatic Sciences
    • /
    • 제1권1호
    • /
    • pp.48-62
    • /
    • 1998
  • Seasonal volume transport on the Texas-Louisiana continental shelf is investigated in terms of objectively fitted transport streamfunction fields based on the current meter data of the Texas­Louisiana Shelf Circulation and Transport Processes Study. Adopted here for the objective mapping is a method employing a two-dimensional truncated Fourier representation of the streamfunction over a domain, with the amplitudes determined by least square fit of the observation. The fitting was done with depth-averaged flow rather than depth-integrated flow to reduce the root-mean-square error. The fitting process filters out $11\%$ of the kinetic energy in the monthly mean transport fields. The shelf-wide pattern of streamfunction fields is similar to that of near-surface velocity fields over the region. The nearshore transport, about 0.1 to 0.3 Sv $(1 Sv= 10^6\;m^3/sec)$, is well correlated with the seasonal signal of along-shelf wind stress. The spring transport is weak compared to other seasons in the inner shelf region. The transport along the shelf break is large and variable. In the southwestern shelf break, transport amounts up to 4.7 Sv, which is associated with the activities of the encroaching of energetic anticyclonic eddies originated in Loop Current of the eastern Gulf of Mexico. The first empirical orthogonal function (EOF) of streamfunction variability contains $67.3\%$ of the variance and shows a simple, shelf-wide, along-shelf pattern of transport. The amplitude evolution of the first EOF is highly correlated (correlation coefficient: 0.88) with the evolution of the along-shelf wind stress. This provides strong evidence that the large portion of seasonal variation of the shelf transport is wind-forced. The second EOF contains $23.7\%$ of the variance and shows eddy activities at the southwestern shelf break. The correlation coefficient between the amplitudes of the second EOF and wind stress is 0.42. We assume that this mode is coupled a periodic inner shelf process with a non-periodic eddy process on the shelf break. The third EOF (accounting for $7.2\% of the variance) shows several cell structures near the shelf break associated with the variability of the Loop Current Eddies. The amplitude time series of the third EOF show little correlation with the along-shelf wind.

  • PDF

실버세대의 심리적 안정감에 따른 건강 기능성 섬유 인지 및 선호에 관한 연구 (A study on Recognition and Preference of Healthy and Functional Textile according to Psychological Comfort of the Silver Generation)

  • 서민녕;구영석
    • 한국의류산업학회지
    • /
    • 제16권5호
    • /
    • pp.811-821
    • /
    • 2014
  • The purposes of this study were to investigate difference on recognition, preference, satisfaction and possession of functional textile clothing in psychological comfort groups according to gender and age of the elderly and suggest optimal physical, mental and healthy functional textiles for the elderly. The data was collected from 262 respondents in their age of over 55, who lived in Busan. SPSS 21.0 was used and t-test, cross tabulation analysis, frequency analysis, and descriptive analysis were performed for analysis. The results are as follows. First, high groups of psychological comfort for both gender and age of the elderly showed higher recognition and preference of functional textiles, especially in the high groups of the new silver generation and women. Second, the high groups of psychological comfort for both gender and age of the elderly showed higher satisfaction of functional textiles. However, it was lower than preference of functional textiles. Last, the possession of major clothing function over all the elderly was comfort, fitting motion adaptability, and health safety function in order. Low groups of psychological comfort possessed fitting motion adaptability textile clothing more than the high groups, and the high groups of psychological comfort possessed comfort and health safety textile clothing more than low groups. In conclusion, this research showed the importance of functional textiles for the elderly in terms of psychological comfort as well as the need for healthy comfort textiles for the advanced activities of daily life.

상호 노드 정보를 이용한 클러스터 기반 퍼지 모델트리 (Cluster Based Fuzzy Model Tree Using Node Information)

  • 박진일;이대종;김용삼;조영임;전명근
    • 한국지능시스템학회논문지
    • /
    • 제18권1호
    • /
    • pp.41-47
    • /
    • 2008
  • 클러스터 기반 퍼지 모델트리에서 훈련 데이터의 과잉 적응은 검중 데이터의 성능을 저하시키는 문제점을 가지고 있다. 이러한 문제점을 해결하기 위한 방법으로 본 논문에서는 상호 노드간의 정보를 고려하는 방법을 제안하고자 한다. 제안된 방법은 우선 입력과 출력변수의 속성을 고려한 퍼지 클러스터링에 의해 중심벡터를 계산한 후, 중심벡터들과 입력 속성간의 소속도를 이용하여 구간 분할된 영역별로 각각의 선형모델을 구축한다. 예측 단계에서는 입력된 데이터가 잎노드에 도달하기까지 경유하게 되는 노드들의 중심벡터들과 입력 데이터간의 거리값에 따른 소속도를 계산한 후 최종적으로 각 노드의 선형모델들과 계산된 소속도를 이용하여 출력값을 예측하게 된다. 제안된 방법의 우수성을 보이기 위해 다양한 벤치마크 데이터를 대상을 실험한 결과, 기존의 클러스터 기반 퍼지 모델트리보다 향상된 성능을 보임을 알 수 있었다.

저수지 유입량 예측을 위한 신경망 모형의 특성 연구 (A Study on Characteristics of Neural Network Model for Reservoir Inflow Forecasting)

  • 김재형;윤용남
    • 한국방재학회 논문집
    • /
    • 제2권4호
    • /
    • pp.123-129
    • /
    • 2002
  • 본 연구에서는 3층 신경망 모형에 의해 충주호의 유입량을 예측한 결과들을 이용하여 신경망 모형의 저수지 유입량 예측 특성을 분석하였다. 신경망 모형의 적절한 입력층 및 은닉층 뉴런 개수, 학습회수를 제시하였으며, 학습 첨두유량 크기가 예측된 첨두유량보다 작을 경우 예측 값이 과소평가되는 특징을 확인하였다. 또한 뉴런 개수, 학습회수가 과다할 경우 발생 가능한 과적합 현상을 확인하였으며, 정확한 예측을 위해 필요한 최소 학습자료 기간도 제시하였다. 결과적으로 충주호의 경우 $8{\sim}10$개의 뉴런 개수 및 $1500{\sim}3000$회의 학습회수를 이용한 신경망 모형이 적합한 것으로, 학습자료 기간 수는 최소한 600개 이상의 자료를 적용하여야 정확한 예측이 가능한 것으로 결과되었다.

네트워크 비정상 탐지를 위한 속성 축소를 반영한 의사결정나무 기술 (Decision Tree Techniques with Feature Reduction for Network Anomaly Detection)

  • 강구홍
    • 정보보호학회논문지
    • /
    • 제29권4호
    • /
    • pp.795-805
    • /
    • 2019
  • 최근 알려지지 않은 공격에 대처하기 위한 네트워크 비정상(anomaly) 탐지 기술에 대한 관심이 한층 높아지고 있다. 이러한 기술 개발을 위해 데이터 마이닝(data mining), 기계학습(machine learning), 그리고 딥러닝(deep learning)등을 활용한 다양한 연구가 진행되고 있다. 본 논문에서는 분류(classification) 문제를 다루는 데이터 마이닝 기술 중 가장 전통적인 방법 중 하나인 의사결정나무(decision tree)를 이용하여 NSL-KDD 데이터 셋을 대상으로 네트워크 비정상 탐지 가능성을 보여준다. 의사결정나무의 과대적합(over-fitting) 단점을 해소하기 위해 카이-제곱(chi-square) 테스트를 통해 최적의 속성 선택(feature selection)을 수행하고, 선택된 13개의 속성을 사용한 의사결정나무 모델 환경에서 NSL-KDD 시험 데이터 셋 KDDTest+에 대해 84% 그리고 KDDTest-21에 대해 70%의 네트워크 비정상 검출 정확도를 보였다. 제시된 정확도는 기존 의사결정나무 모델 적용 시 이들 시험 데이터 셋을 대상으로 알려진 정확도 81% 그리고 64% 수준과 비교해 약 3% 그리고 6% 각각 향상된 결과다.

의사결정나무 기법을 이용한 노인들의 자살생각 예측모형 및 의사결정 규칙 개발 (A Development of Suicidal Ideation Prediction Model and Decision Rules for the Elderly: Decision Tree Approach)

  • 김덕현;유동희;정대율
    • 한국정보시스템학회지:정보시스템연구
    • /
    • 제28권3호
    • /
    • pp.249-276
    • /
    • 2019
  • Purpose The purpose of this study is to develop a prediction model and decision rules for the elderly's suicidal ideation based on the Korean Welfare Panel survey data. By utilizing this data, we obtained many decision rules to predict the elderly's suicide ideation. Design/methodology/approach This study used classification analysis to derive decision rules to predict on the basis of decision tree technique. Weka 3.8 is used as the data mining tool in this study. The decision tree algorithm uses J48, also known as C4.5. In addition, 66.6% of the total data was divided into learning data and verification data. We considered all possible variables based on previous studies in predicting suicidal ideation of the elderly. Finally, 99 variables including the target variable were used. Classification analysis was performed by introducing sampling technique through backward elimination and data balancing. Findings As a result, there were significant differences between the data sets. The selected data sets have different, various decision tree and several rules. Based on the decision tree method, we derived the rules for suicide prevention. The decision tree derives not only the rules for the suicidal ideation of the depressed group, but also the rules for the suicidal ideation of the non-depressed group. In addition, in developing the predictive model, the problem of over-fitting due to the data imbalance phenomenon was directly identified through the application of data balancing. We could conclude that it is necessary to balance the data on the target variables in order to perform the correct classification analysis without over-fitting. In addition, although data balancing is applied, it is shown that performance is not inferior in prediction rate when compared with a biased prediction model.

Optimized Geometric LDPC Codes with Quasi-Cyclic Structure

  • Jiang, Xueqin;Lee, Moon Ho;Gao, Shangce;Wu, Yun
    • Journal of Communications and Networks
    • /
    • 제16권3호
    • /
    • pp.249-257
    • /
    • 2014
  • This paper presents methods to the construction of regular and irregular low-density parity-check (LDPC) codes based on Euclidean geometries over the Galois field. Codes constructed by these methods have quasi-cyclic (QC) structure and large girth. By decomposing hyperplanes in Euclidean geometry, the proposed irregular LDPC codes have flexible column/row weights. Therefore, the degree distributions of proposed irregular LDPC codes can be optimized by technologies like the curve fitting in the extrinsic information transfer (EXIT) charts. Simulation results show that the proposed codes perform very well with an iterative decoding over the AWGN channel.

자율이동로봇의 안전주행을 위한 주행성 맵 작성 (Development of a Traversability Map for Safe Navigation of Autonomous Mobile Robots)

  • 진강규
    • 제어로봇시스템학회논문지
    • /
    • 제20권4호
    • /
    • pp.449-455
    • /
    • 2014
  • This paper presents a method for developing a TM (Traversability Map) from a DTM (Digital Terrain Model) collected by remote sensors of autonomous mobile robots. Such a map can be used to plan traversable paths and estimate navigation speed quantitatively in real time for robots capable of performing autonomous tasks over rough terrain environments. The proposed method consists of three parts: a DTM partition module which divides the DTM into equally spaced patches, a terrain information module which extracts the slope and roughness of the partitioned patches using the curve fitting and the fractal-based triangular prism method, and a traversability analysis module which assesses traversability incorporating with extracted terrain information and fuzzy inference to construct a TM. The potential of the proposed method is validated via simulation works over a set of fractal DTMs.