• 문화기술의 융합
• /
• 제10권3호
• /
• pp.461-466
• /
• 2024

한국의 주민참여예산제도는 자치단체별로 자율적으로 운영되도록 하고 있어서, 본 연구는 이들을 몇 개의 유사한 유형들로 구분하여서 각각의 특징들을 살펴보고자 한다. 본 연구는 다양한 머신 러닝 기법들을 활용하여 2022년도 기초 시(市)를 중심으로 운영유형을 분류하였다. 그 결과, 여러 머신 러닝 기법(Neural Network, Rule Induction(CN2), KNN, Decision Tree, Random Forest, Gradient Boosting, SVM, Naïve Bayes) 중에서 SVM 기법이 성능이 가장 좋은 것으로 확인되었다. SVM 기법이 밝혀낸 운영유형은 모두 3개인데, 하나는 위원회 활동은 적게 하지만, 참여예산은 많이 확보하는 클러스터(C1)이고, 다른 하나는 주민참여예산제에 매우 소극적인 도시들의 클러스터(C3)이다. 마지막 클러스터(C2)는 참여예산에 전반적으로 적극적인데, 대다수 지역이 여기에 해당한다. 결론적으로 한국의 대다수 자치단체는 주민참여예산제를 긍정적으로 운영하고 있으며, 오직 소수의 자치단체만 소극적이다. 후속 연구로 지난 10여 년간의 시계열 자료를 분석한다면, 우리는 주민참여예산에 관한 지방자치단체 유형 분류의 신뢰도를 더욱 높일 수 있을 것으로 기대한다.

• 한국재료학회지
• /
• 제5권7호
• /
• pp.869-879
• /
• 1995

TiAi intermetallic compound has been extensively studied for possible high temperature structural applications because of its high specific strength at high temperature, high creep resistance, and good oxidation resistance at elevated temperatures. In addition to its good properties, an economic manufacturing routes should be developed for this material to be used more extensively. One of the promising route in manufacturing TiAl intermetallics is the Self-propagating High-temperature Synthesis (SHS) method. Thus in this study, an attempt was made to study the mechanism of the SHS process in TiAl synthesis. The composition of the sample was Ti-(45, 50, 53)at% Al and the microstuctures of the products were analyzed using optical microscope and scanning electron microscope. When the phases formed at the main SHS reaction of whicyh combustion temperature is higher than the melting temperature of aluminum were identified as TiAl and Ti$_3$Al ; Ti$_3$Al cores surrounded by TiAl phase. In order to increase the combustion temperature, carbon was added 5 and 10at.%. When the carbon content was 10at.%, the heat of the reaction was large enough to melt the phase formed and that is consistent with the theoretical calculation results of the adiabatic temperature. The combution temperatue, which was measured by a computer data acquisition system, increased with the carbon content. The phases formed from the reaction involving the carbon added were indentified as TiAl and Ti$_2$AlC using XRD. The vickers hardness of the reaction product increased with the carbon content.

• 한국식품과학회지
• /
• 제19권2호
• /
• pp.119-124
• /
• 1987

딸기쥬스와 오렌지쥬스 농축액의 빙점을 ${-15}^{\circ}C$로 강하시킬수 있는 복합cryoprotectants를 선정하고, 이를 쥬스에 첨가하여 액체상태로 $-15^{\circ}C$에서 저장하면서 품질변화를 조사하였다. 딸기쥬스에 적합한 복합 cryprotectants의 조성은 설탕25.5%(w/w), 포도당 12.7%, glycerol 1%, propylene glycol 1% 및 ascorbic acid 0.1%였다. 그리고 48% 고형분을 함유하는 농축 오렌지쥬스에 적합한 것으로는 포도당 5%, 과당 5%, glycerol 4% 및 citricacid 1%였다. 저장중 과실쥬스의 종합적 품질은 복합 cryprotectants를 첨가한 것과 대조구사이에 차이가 없었고, 4개월 저장한 딸기쥬스로만들은쨈의 품질에 있어서도 대조구와 처리구간에 유의성 있는 품질차이가 없었다. 이들 결과로 보아 과실쥬스에 복합 cryoprotectants를 첨가하여 냉통온도에서 액체상태로 저장하면 보통 방법으로 냉동저장한것과 대등한 품질을 보존할수 있는 것으로 밝혀졌다.

• 한국지반신소재학회논문집
• /
• 제19권1호
• /
• pp.45-53
• /
• 2020

철도분야에서도 계측자료를 바탕으로 머신러닝 기법을 이용하여 예측 분석하는 시도가 점차적으로 증가하고 있는 실정이다. 이 논문에서는 열차의 차상가속도 데이터를 기반으로 궤도의 품질을 결정하는 지표 중에 하나인 궤도품질지수를 머신러닝 기법을 활용하여 예측하였다. 머신러닝 기법으로 활용하고 있는 대표적인 3개의 모델로 궤도품질지수를 예측하여 가장 정확도가 높은 모델은 XGBoost으로 데이터셋에서 85% 이상의 예측정확도를 보였다. 또한 윤축과 대차의 z축의 진동가속도가 고저 궤도품질지수의 기여도가 높은 것으로 나타났으며, 이는 기존 연구결과와도 잘 일치하였다. 이러한 결과를 볼 때 단일 알고리즘인 서포터 벡터머신보다는 앙상블 알고리즘을 적용한 랜덤포레스트와 XGBoost이 정확도가 높은 것으로 판단된다. 따라서 머신러닝 기법에서 적용모델에 따라 정확도가 달라질 수 있기 때문에 차량진동가속도를 이용한 궤도품질지수를 예측하기 위해서는 앙상블 알고리즘을 가지는 모델을 적용하는 것이 적절할 것으로 판단된다.

• 한국표면공학회:학술대회논문집
• /
• 한국표면공학회 2001년도 춘계학술발표회 초록집
• /
• pp.3-3
• /
• 2001

Recently TiN, TiAlN, CrN hardcoatings have adapted many industrial application such as die, mold and cutting tools because of good wear resistant and thermal stability. However, in terms of high speed process, general hard coatings have been limited by oxidation and thermal hardness drop. Especially in the case of PCB drill, high speed cutting and without lubricant process condition have not adapted these coatings until now. Therefore more recently, superhard nanocomposite coating which have superhard and good thermal stability have developed. In previous works, WC-TiAlN new nanocomposite film was investigated by cathodic arc ion plating system. Control of AI concentration, WC-TiAlN multi layer composite coating with controlled microstructure was carried out and provides additional enhancement of mechanical properties as well as oxidation resistance at elevated temperature. It is noted that microhardness ofWC-TiA1N multi layer composite coating increased up to 50 Gpa and got thermal stability about $900^{\circ}C$. In this study WC-TiAlN nanocomposite coating was deposited on PCB drill for enhancement of life time. The parameter was A1 concentration and plasma cleaning time for edge sharpness maintaining. The characteristic of WC-TiAlN film formation and wear behaviors are discussed with data from AlES, XRD, EDS and SEM analysis. Through field test, enhancement of life time for PCB drill was measured.

• 대한전자공학회논문지SD
• /
• 제42권1호
• /
• pp.19-24
• /
• 2005

본 논문에서는 압저항 효과를 이용한 Si 압력센서 제작을 최적화하였다. Si 압저항형 압력센서의 제작공정에 있어서 압저항과 알루미늄 회로 패턴 이후에 Si 이방성 식각을 통하여 수율이 개선되었다. 압저항의 위치와 공정 파라메터는 각각 ANSYS와 SUPREME 시뮬레이터를 이용하여 결정하였다. Boron-depth 프로파일 측정으로부터 p-형 Si 압저항의 두께를 측정한 결과 SUPREME 시뮬레이션으로부터 얻은 결과와 잘 부합하였다. 다이아프램을 위한 Si 이방성 식각 공정은 암모늄 첨가제 AP(Ammonium persulfate)를 TMAH(Tetra-methyl ammonium hydroxide) 용액에 첨가함으로써 최적화되었다.

• 대한기계학회논문집B
• /
• 제30권8호
• /
• pp.772-779
• /
• 2006

Window is a critical component in the design of energy-efficient buildings. To minimize the heat loss, insulation performance of the glazing has to be improved. Manufacturing of vacuum glazing has been motivated by the possibility of making windows of very good thermal insulation properties for such applications. It is made by maintaining vacuum in the gap between two glass panes. Pillars are placed between them to withstand the atmospheric pressure. Edge covers are applied to reduce conduction through the edge. Accurate measurements have been made of the radiative heat transfer, the pillar conduction and the gas conduction using a guarded hot plate apparatus. Vacuum glazing is found to have low thermal conductance roughly below $1W/m^2K$. Among the heat transfer modes of residual gas conduction, conduction through support pillar and the radiative heat transfer between the glass panes, the last one is the most dominant to the overall thermal conductance. Vacuum glazing using very low emittance AI-coated glass has an overall thermal conductance of about $0.7W/m^2K$.

• 한국경영과학회지
• /
• 제22권3호
• /
• pp.37-58
• /
• 1997

An AI approach with tabu search is designed to solve multi-level knapsack problems. The approach performs intelligent actions with memories of historic data and learning effect. These action are developed ont only by observing the attributes of the optimal solution, the solution space, and its corresponding path to the optimal, but also by applying human intelligence, experience, and intuition with respect to the search strategies. The approach intensifies, or diversifies the search process appropriately in time and space. In order to create a good neighborhood structure, this approach uses two powerful choice rules that emphasize the impact of candidate variables on the current solution with respect to their profit contribution. "Pseudo moves", similar to "aspirations", support these choice rules during the evaluation process. For the purpose of visiting as many relevant points as possible, strategic oscillation between feasible and infeasible solutions around the boundary is applied. To avoid redundant moves, short-term (tabu-lists), intemediate-term (cycle-detection), and long-term (recording frequency and significant solutions for diversfication) memories are used. Test results show that among the 45 generated problems (these problems pose significant or insurmountable challenges to exact methods) the approach produces the optimal solutions in 39 cases.lutions in 39 cases.

• 한국방사성폐기물학회:학술대회논문집
• /
• 한국방사성폐기물학회 2004년도 Proceedings of the 4th Korea-China Joint Workshop on Nuclear Waste Management
• /
• pp.73-82
• /
• 2004

This paper studies the leaching behaviors of pyrochlore-rich synroc incorporated 46.8wt% simulated actinides waste under the five simulated geological disposal media, which included the bentonite, granite, granite ＋ ferroferric oxide, granite ＋ cement, bentonite ＋ ferroferric oxide, respectively. The mass loss rates reached to equilibrium after 182 day and was 10-7 g/$\textrm{mm}^2{\cdot}d$. That suggests the mass loss rate of pyrochlore-rich synroc, loaded 46.8wt% actinides waste, was very low. The surfaces of the leached specimens were analyzed by XRD, SEM/EDS. The experimental results show that the pyrochlore-rich synroc samples in the systems, which contained bentonite and cement, have two new phases formed on the leached specimens surface at $90^{\circ}C$ for 728d; The bentonite and cement can retard the elements leaching; $Fe_3O_4$ can speed the elements leaching; Expect for Ti ion depleted on the sample surface, other ion, such as U, Zr, AI, Ca, were in equable states and Ba ion was enriched during test time, which indicated the simulated disposal media have good ability to retard the leaching behavior of the pyrochlore-rich synroc.

• 한국인공지능학회지
• /
• 제6권2호
• /
• pp.11-15
• /
• 2018

Breast cancer is one of the leading causes of cancer related death among women. So prediction of overall survival status is important into decided in adjuvant treatment. Deep belief network is a kind of artificial intelligence (AI). We intended to construct prediction model by deep belief network using associated clinicopathologic factors. 103881 cases were found in the Korean Breast Cancer Registry. After preprocessing of data, a total of 15733 cases were enrolled in this study. The median follow-up period was 82.4 months. In univariate analysis for overall survival (OS), the patients with advanced AJCC stage showed relatively high HR (HR=1.216 95% CI: 0.011-289.331, p=0.001). Based on results of univariate and multivariate analysis, input variables for learning model included 17 variables associated with overall survival rate. output was presented in one of two states: event or cencored. Individual sensitivity of training set and test set for predicting overall survival status were 89.6% and 91.2% respectively. And specificity of that were 49.4% and 48.9% respectively. So the accuracy of our study for predicting overall survival status was 82.78%. Prediction model based on Deep belief network appears to be effective in predicting overall survival status and, in particular, is expected to be applicable to decide on adjuvant treatment after surgical treatment.