• Geomechanics and Engineering
• /
• 제25권1호
• /
• pp.1-16
• /
• 2021

Machine learning models have been widely used for landslide susceptibility assessment (LSA) in recent years. The large number of inputs or conditioning factors for these models, however, can reduce the computation efficiency and increase the difficulty in collecting data. Feature selection is a good tool to address this problem by selecting the most important features among all factors to reduce the size of the input variables. However, two important questions need to be solved: (1) how do feature selection methods affect the performance of machine learning models? and (2) which feature selection method is the most suitable for a given machine learning model? This paper aims to address these two questions by comparing the predictive performance of 13 feature selection-based machine learning (FS-ML) models and 5 ordinary machine learning models on LSA. First, five commonly used machine learning models (i.e., logistic regression, support vector machine, artificial neural network, Gaussian process and random forest) and six typical feature selection methods in the literature are adopted to constitute the proposed models. Then, fifteen conditioning factors are chosen as input variables and 1,017 landslides are used as recorded data. Next, feature selection methods are used to obtain the importance of the conditioning factors to create feature subsets, based on which 13 FS-ML models are constructed. For each of the machine learning models, a best optimized FS-ML model is selected according to the area under curve value. Finally, five optimal FS-ML models are obtained and applied to the LSA of the studied area. The predictive abilities of the FS-ML models on LSA are verified and compared through the receive operating characteristic curve and statistical indicators such as sensitivity, specificity and accuracy. The results showed that different feature selection methods have different effects on the performance of LSA machine learning models. FS-ML models generally outperform the ordinary machine learning models. The best FS-ML model is the recursive feature elimination (RFE) optimized RF, and RFE is an optimal method for feature selection.

• 한국의류학회지
• /
• 제37권4호
• /
• pp.554-564
• /
• 2013

This study identified the characteristic and meaning of flow experience through the experiences related to fashion models' role performance; subsequently, the following findings were obtained. First, fashion models' flow experience could be divided into characteristics such as temporality, spatiality, relationality, physicalness and pleasure. Second, the process of this flow experience ultimately led to the complete moment. The complete moment can be said to be the aesthetic experience that provides both the meaningful experience and the aesthetic pleasure; it is the experiential knowledge at the dimension of mysterious integration that their body and mind are integrated into one. The beauty that fashion models exercise at the aspect of this aesthetic experience is that of performance and is an individual physical movement where they perceive their role and exercise their inner ability to express their costume most beautifully. Accordingly, the beauty of performance can be said to mean that the fashion show was successfully held by inducing both the performer and the audience into an aesthetic response. The process of specialized planning and preparation is required for fashion models to exercise the beauty of performance at the complete movement reached through flow experience and a successful fashion show. Diverse elements of the fashion show should be more organically constituted through such a process. Fashion models should exert efforts to embody acts such as walking, posing and turning through the performance of their excellent role as well as develop a training program to complete it.

• 대한기계학회논문집
• /
• 제14권1호
• /
• pp.241-250
• /
• 1990

본 연구에서는 낙구식 점도계를 이용하여 낮은 농도의 용액을 실험할 때 종말 속도측정의 어려움을 해결하기 위하여 레이저와 특수 타이머를 설치하였으며 정확한 특성시간을 결정하기 위하여 실린더 내부의 시험유체를 교란시키지 않고, 떨어뜨린 구 를 회수하는 장치를 제작하였다. 또한 주로 rheogoniometer에 의존하던 영 전단률 점성계수를 측정하기 위하여 속이 빈 알루미늄 구(hollow aluminium ball)의 밀도를 시험 유체와 거의 같은 정도까지 변화시켜가며 종말 속도를 측정하였으며 점탄성 유체 로써 Separan AP-273의 낮은 농도인 300에서 2000wppm까지의 저농도 용액에 대한 특성 시간을 여러모델에 의하여 실험적으로 결정하고 저농도 폴리머 용액에서 퇴화로 인한 점탄성유체의 특성 변화를 분석하였다.

• 한국지반환경공학회 논문집
• /
• 제19권10호
• /
• pp.21-26
• /
• 2018

흙-습윤 특성곡선은 불포화토에서 물의 흐름, 다상유체에서의 상대투수계수, 그리고 흙의 강성 및 부피 변화를 이해하는데 필수적이다. 바이오폴리머는 미생물에 완전 분해가 가능한 자연에 무해한 친환경 물질이다. 따라서, 치토산, 폴리에틸렌 옥사이드, 잔탄검, 알지네이트염 및 폴리아크릴산 등과 같은 바이오폴리머가 지반복원, 지반성능향상 및 오일생산증진을 위해 연구되어왔다. 폴리아크릴산(polyacrylic acid)은 바이오폴리머의 일종으로, 유체의 흐름 특성 향상을 통하여 오일생산증진 및 지반복원 등의 분야에서 우수한 성능을 보여왔다. 따라서, 본 연구에서는 폴리아크릴산(polyacrylic acid)을 포함한 모래의 흙-습윤 특성곡선의 이해하기 위해서 실내시험을 수행하고, 이론적 모델의 매개변수 변화를 연구하였다. 그 결과, 폴리아크릴산(polyacrylic acid)의 농도가 증가함에 따라, 포화된 모래에 주입되는 공기의 주입 모세관압력이 증가함을 보이고, 높은 모세관압력에서의 잔류 함수비 역시 증가하고 있음을 보였다. 또한, 이론식 모델의 매개변수를 구하였으며, 이를 활용한 흙-습윤 특성곡선이 실내 실험 결과와 잘 일치하고 있음을 보였다. 따라서, 이론적 모델의 매개변수를 알고 있으면, 본 연구에서 활용된 폴리아크릴산(polyacrylic acid)이외의 바이오폴리머를 포함한 모래 지반의 흙-습윤 특성곡선이 예측이 가능함을 설명하였다.

• 대한기계학회:학술대회논문집
• /
• 대한기계학회 2007년도 춘계학술대회A
• /
• pp.1469-1474
• /
• 2007

In recent years, research to machine large-surface micro-features has become important because of the light guide panel of a large-scale liquid crystal display and the bipolar plate of a high-capacity proton exchange membrane fuel cell. In this study, in order to realize the systematic design technology and performance improvements of an ultra-precision machine for machining the large-surface micro-features, a structural characteristic analysis was performed using its virtual prototype. The prototype consisted of gantry-type frame, hydrostatic feed mechanisms, linear motors, brushless DC servo motor, counterbalance mechanism, and so on. The loop stiffness was estimated from the relative displacement between the tool post and C-axis table, which was caused by a cutting force. Especially, the causes of structural stiffness deterioration were identified through the structural deformation analysis of sub-models.

• 한국감성과학회:학술대회논문집
• /
• 한국감성과학회 2000년도 춘계 학술대회 및 국제 감성공학 심포지움 논문집 Proceeding of the 2000 Spring Conference of KOSES and International Sensibility Ergonomics Symposium
• /
• pp.157-162
• /
• 2000

Discriminant analysis is a method to relate whether the objects have a specific characteristic or not with their 'continuous' attribute values and, for given objects, to estimate whether they have a specific characteristic or not by their values of discriminant scores gotten from their attribute values. The author developed the new 'computational' method of discriminant analysis without specific hypotheses or assumptions and, by this new method, we can find 'feasible' solutions under the conditions required by our actual problems. In this paper, the author tried to apply this new method to the discrimination of logical relations. If this trial could be a success, we can apply this new method of discriminant analysis to the problems about relating the specific characteristic of the objects with their 'discrete' attribute values. The result was successful and the applicability of discriminant analysis could be expanded as a method for constructing the models for "estimating impressions".

• 품질경영학회지
• /
• 제23권3호
• /
• pp.20-32
• /
• 1995

This paper is concerned with the economic selection of both the lower limit and the process mean for a continuous production process. Consider a production process where items are produced continuously. All of the items are subject to acceptance inspection. The items for which the measured values of the quality characteristic are larger than the lower limit are accepted, and those smaller than the lower limit are rejected and excluded from shipment. The process mean may be set higher to reduce the costs incurred by imperfect quality. Using a higher process mean, however, results in a higher production cost when production cost is an increasing function of the quality characteristic. Assuming that the quality characteristic is normally distributed with known variability, cost models are constructed which involve production cost, cost incurred by imperfect quality, rejection cost, and inspection cost. Methods of finding optimal values of the lower limit and the process mean are presented and numerical examples are given.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2015년도 제46회 하계학술대회
• /
• pp.756-757
• /
• 2015

This paper deals with characteristic analysis of surface mounted permanent magnet synchronous motor according to a phase-separation of stator. The characteristic analysis of designed model is performed by finite element analysis(FEA), and the result are shown that there is no difference between the two models in comparison with a general SPMSM. Finally, this study verifies the feasibility of SPMSM with separated stator core. The experimental data is shown the validity of this paper.

• Journal of Advanced Marine Engineering and Technology
• /
• 제28권8호
• /
• pp.1279-1285
• /
• 2004

This paper presents the characteristic simulation of PM-type magnetic circuit breaker with the 2D finite element magnetic field solution including non-linearity of the material and an eddy current. Change of dynamic characteristic of the actuator is quantified from the finite element analysis. The results obtained from a commercial finite element analysis software are compared with those calculated from the developed finite element analysis software. A new modified model to decrease the eddy current is proposed. The characteristics of the two models are compared.

• 소음진동
• /
• 제7권5호
• /
• pp.781-787
• /
• 1997

A new method is presented to determine two fundamental acoustic quantities of sound-absorbing materials such as characteristic impedance and propagation constant. In this study, the surface acoustic impedances of sound-absorbing materials are measured using the impedance tube and the anechoic chamber to determine the above acoustic quantities. The measured results are given for two typical sound-absorbing materials(glass wool and urethane foam) int the frequency range between 150 and 1, 600 Hz. The results are verified by other two known methods, which are Smith & Parrott method and Utsuno et al. method. The absorption coefficients calculated from the empirical models(Miki model for glass wool and Jung model for urethane foam) and two quantities by present method are in good agreement with the measured values.