• Journal of the Korean Data and Information Science Society
• /
• v.24 no.1
• /
• pp.125-133
• /
• 2013

In this paper, penalized binary logistic regression models are employed as statistical models for determining the discharge of 668 patients with a chief complaint of dyspnea based on 11 blood tests results. Specifically, the ridge model based on $L^2$ penalty and the Lasso model based on $L^1$ penalty are considered in this paper. In the comparison of prediction accuracy, our models are compared with the logistic regression models with all 11 explanatory variables and the selected variables by variable selection method. The results show that the prediction accuracy of the ridge logistic regression model is the best among 4 models based on 10-fold cross-validation.

• The Bulletin of The Korean Astronomical Society
• /
• v.37 no.2
• /
• pp.176.2-176.2
• /
• 2012

다목적실용위성 3호의 태양전지판은 위성의 -Z축 방향에 고정되어 있는 방식으로 사용되고 있다. 이로 인해 위성이 임무수행을 위한 자세기동을 하게 되면 태양전지판의 태양입사각 변화에 따라 전력생산량이 변하게 되고 이를 예측하여 최대 방전률(DOD : Depth of Discharge)을 넘지 않는 제한조건 내에서 임무 계획을 수행해야 한다. 전력생산량 및 전력소비량을 예측하기 위해서는 전력 모의실험을 수행해야 하며 이를 위해 위성의 자세 및 위치정보, 임무를 고려한 Mission Profile, 태양입사각, 초기 방전률 값을 생성해야 한다. 본 논문은 태양입사각 계산을 위해 위성의 임무(영상 촬영, 지상국 교신)를 반영한 자세 및 위치 정보를 생성하고, 이 결과를 태양입사각 계산 로직에 적용하여 태양입사각을 생성한 결과를 정리하였다. 생성된 결과의 타당성을 검토하기 위해 상용 툴인 STK를 이용하여 비교를 수행하였다. 또한, 전력 모의실험에 사용된 Mission Profile은 위성 운용에 안정성을 높이며 복잡한 임무 시나리오에 적용이 용이하도록 운용 Margin을 고려하여 생성하였다. 본 논문에서 제시한 방안을 실제 수행된 임무 시나리오에 적용하여 전력 모의실험을 수행하였으며, 그 결과를 임무 수행 후 획득된 위성 Telemetry를 이용한 실측값과 비교하여 전력 모의 실험 결과에 대한 타당성을 검증하였다. 실제 초기 운영결과 제한된 전력 허용 범위 내에서 적용이 가능함을 확인할 수 있었다.

• Journal of KIISE:Computer Systems and Theory
• /
• v.27 no.8
• /
• pp.741-751
• /
• 2000

To achieve high performance by exploiting instruction level parallelism aggressively in superscalar processors, it is necessary to overcome the limitation imposed by control dependences and data dependences which prevent instructions from executing parallel. Value prediction is a technique that breaks data dependences by predicting the outcome of an instruction and executes speculatively its data dependent instruction based on the predicted outcome. In this paper, a hybrid value prediction scheme with dynamic classification mechanism is proposed. We design a hybrid predictor by combining the last predictor, a stride predictor and a two-level predictor. The choice of a predictor for each instruction is determined by a dynamic classification mechanism. This makes each predictor utilized more efficiently than the hybrid predictor without dynamic classification mechanism. To show performance improvements of our scheme, we simulate the SPECint95 benchmark set by using execution-driven simulator. The results show that our scheme effect reduce of 45% hardware cost and 16% prediction accuracy improvements comparing with the conventional hybrid prediction scheme and two-level value prediction scheme.

• Aerospace Engineering and Technology
• /
• v.7 no.2
• /
• pp.131-143
• /
• 2008

Slag mass deposition was required to predict performance accurately of KSLV-I kick motor(KM) system. The validation of the numerical analysis was performed with mass flow rate measured at 4th ground test of the KM. The study described here included internal flow field of KM at various time steps during burning. Slag mass accumulation was computed through the aluminum oxide particle paths to deviate from the gas flow streamlines in flight. These numerical analysis was performed with Fluent 6.3 program The effects for the acceleration, origins and diameters of the aluminum oxide particles was analyzed, finally the total slag mass accumulation was acquired. We confirmed that the slag mass deposition was agreement well with predicted slag mass based on kick motor the grounded test.

• Geotechnical Engineering
• /
• v.14 no.6
• /
• pp.127-138
• /
• 1998

In refuse landfills, a considerable amount of settlement occurs due to the decomposition of refuse over several years. In this paper, several prediction methods are applied to the measured settlement data of fresh refuse sites. The effect of biological decomposition on the settlement characteristics is investigated in predicting the long-term settlement of refuse landfill sites in view of the predicted settlement curves and the amount of long-term settlement. Irrespective of the applied models, the long term settlement may not be correctly estimated if the model parameters do not contain the decomposition effects. Among the proposed several prediction methods, Gibson & Lo model and hyperbolic model seem to represent the long-term settlement characteristics, but the power creep law seems to considerably overestimate the long-term settlement.

• Geotechnical Engineering
• /
• v.13 no.2
• /
• pp.77-90
• /
• 1997

A resonable and economical method which can predict permeability of rock mass in underground is needed to overcome the uncertainty of groundwater behavior. For this par pose, one prediction method of permeability has been studied. The artificial neural networks model using error back propagation algorithm, . one of the teaching techniques, is utilized for this purpose. In order to verify the applicability of this model, in-situ permeability results are simulated. The simulation results show the potentiality of utilizing the neural networks for effective permeability prediction of rock mass.

• Journal of the Korean Data and Information Science Society
• /
• v.28 no.3
• /
• pp.635-645
• /
• 2017

In this paper, we propose batting index prediction models of 2017. Due to the insufficiency of KBO pitchers data, batting index prediction models of 2016 has been developed based on elected eight batting index collecting the past three years data of MLB and KBO. It has been found that this prediction model fits well to both MLB and KBO, and the KBO model fits better than MLB in some cases. Using these prediction models, we analyzed and compared 2016's estimated values for the batting index of MLB and KBO. With the relation results between batting index prediction and batter's age for MLB and KBO, it can be determined that there is no relationship between the significant batting index and ages.

• Journal of KIISE:Computer Systems and Theory
• /
• v.28 no.11
• /
• pp.601-612
• /
• 2001

In superscalar processors, value prediction is a technique that breaks true data dependences by predicting the outcome of an instruction in order to exploit instruction level parallelism(ILP). A value predictor looks up the prediction table for the prediction value of an instruction in the instruction fetch stage, and updates with the prediction result and the resolved value after the execution of the instruction for the next prediction. However, as the instruction fetch and issue rates are increased, the same instruction is likely to fetch again before is has been updated in the predictor. Hence, the predictor looks up the stale value in the table and this mostly will cause incorrect value predictions. In this paper, a stride value predictor with the capability of speculative updates, which can update the prediction table speculatively without waiting until the instruction has been completed, is proposed. Also, the performance of the scheme is examined using Simplescalar simulator for SPECint95 benchmarks in which our value predictor is added.

• Korean Journal of Construction Engineering and Management
• /
• v.24 no.2
• /
• pp.98-110
• /
• 2023

The smart city public and private partnership project is a project to build and operate a sustainable city by investing land and capital in public-private partnership to build urban infrastructure and providing various urban services. It highly depends on the precise feasibility study and the projection of the various factors affecting the project during the planning stage to get the project successful. However, it is very difficult to predict the possibility of the project success in advance due to various physical and social factors. It is necessary to derive factors affecting the project at the planning stage and respond with appropriate analysis in order to solve these problems and to carry out a successful project. Therefore, this study derived preliminary components for feasibility study through previous studies and order status analysis and presented feasibility study components such as five-step processes, 10 process items, 19 analysis items, and 54 detailed analysis items through the Delphi method. It can be expected that this research is to contribute corresponding to diversified possible risks and facilitate the projects during the promotion.

• The Journal of the Acoustical Society of Korea
• /
• v.21 no.6
• /
• pp.522-530
• /
• 2002

This paper introduces a neural network (NN) -based nonlinear predictor for the LP (Linear Prediction) residual. To evaluate the effectiveness of the NN-based nonlinear predictor for LP-residual, we first compared the average prediction gain of the linear long-term predictor with that of the NN-based nonlinear long-term predictor. Then, the effects on the quantization noise of the nonlinear prediction residuals were investigated for the NN-based nonlinear predictor A new NN predictor takes into consideration not only prediction error but also quantization effects. To increase robustness against the quantization noise of the nonlinear prediction residual, a constrained back propagation learning algorithm, which satisfies a Kuhn-Tucker inequality condition is proposed. Experimental results indicate that the prediction gain of the proposed NN predictor was not seriously decreased even when the constrained optimization algorithm was employed.