• Management Science and Financial Engineering
• /
• 제21권2호
• /
• pp.9-11
• /
• 2015

This note is concerned with the use of incomplete weights in multiple criteria decision making. In an earlier study, an optimistic use of incomplete weights is developed to prioritize decision alternatives, which applies the most favorable set of weights to the alternative to be evaluated. In this note, we develop a method for a pessimistic use, thereby applying the least favorable weight set to the evaluated alternative. This development makes possible a more detailed prioritization of competing alternatives, and hence enhances decision-making powers.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제8권4호
• /
• pp.1463-1480
• /
• 2014

It is well known that sparse code is effective for feature extraction of face recognition, especially sparse mode can be learned in the kernel space, and obtain better performance. Some recent algorithms made use of single kernel in the sparse mode, but this didn't make full use of the kernel information. The key issue is how to select the suitable kernel weights, and combine the selected kernels. In this paper, we propose a novel multiple kernel sparse representation based classification for face recognition (MKSRC), which performs sparse code and dictionary learning in the multiple kernel space. Initially, several possible kernels are combined and the sparse coefficient is computed, then the kernel weights can be obtained by the sparse coefficient. Finally convergence makes the kernel weights optimal. The experiments results show that our algorithm outperforms other state-of-the-art algorithms and demonstrate the promising performance of the proposed algorithms.

• 경영과학
• /
• 제19권2호
• /
• pp.139-153
• /
• 2002

The Analytic Hierarchy Process (AHP) is known as a very useful decision-making model developed for obtaining the relative weights of alternatives through pairwise comparison in the context of hierarchical structure. In this paper, we propose a method to reflect the reliability of evaluators in the process of pairwise comparison. This method is applied to the evaluation of aerospace technology alternatives. We have conducted a questionnaire survey for S company that is one of the representative aerospace companies in Korea. A questionnaire was designed for obtaining both the priority with considering the reliability of evaluators' importance weights (the modified AHP priority) and the priority with assuming equally reliable evaluators' importance weights (the AHP priority) in order to compare the priority derived by each of two methods. The result shows that there exists the difference hard to neglect between the final priorities gained by two methods.

• Journal of Multimedia Information System
• /
• 제9권3호
• /
• pp.177-182
• /
• 2022

A deep neural network (DNN) includes variables whose values keep on changing with the training process until it reaches the final point of convergence. These variables are the co-efficient of a polynomial expression to relate to the feature extraction process. In general, DNNs work in multiple 'dimensions' depending upon the number of channels and batches accounted for training. However, after the execution of feature extraction and before entering the SoftMax or other classifier, there is a conversion of features from multiple N-dimensions to a single vector form, where 'N' represents the number of activation channels. This usually happens in a Fully connected layer (FCL) or a dense layer. This reduced 2D feature is the subject of study for our analysis. For this, we have used the FCL, so the trained weights of this FCL will be used for the weight-class correlation analysis. The popular DNN models selected for our study are ResNet-101, VGG-19, and GoogleNet. These models' weights are directly used for fine-tuning (with all trained weights initially transferred) and scratch trained (with no weights transferred). Then the comparison is done by plotting the graph of feature distribution and the final FCL weights.

• 대한수학회지
• /
• 제44권6호
• /
• pp.1429-1440
• /
• 2007

Let ${K^{(\vec{p};N)}_{\vec{n}}(x)}$ be a multiple Kravchuk polynomial with respect to r discrete Kravchuk weights. We first find a lowering operator for multiple Kravchuk polynomials ${K^{(\vec{p};N)}_{\vec{n}}(x)}$ in which the orthogonalizing weights are not involved. Combining the lowering operator and the raising operator by Rodrigues# formula, we find a (r+1)-th order difference equation which has the multiple Kravchuk polynomials ${K^{(\vec{p};N)}_{\vec{n}}(x)}$ as solutions. Lastly we give an explicit difference equation for ${K^{(\vec{p};N)}_{\vec{n}}(x)}$ for the case of r=2.

• 산업경영시스템학회지
• /
• 제26권4호
• /
• pp.31-41
• /
• 2003

This paper deals with a cell formation problem for a set of m-machines and n-processing parts. Generally, a cell formation problem is known as NP-completeness. Hence the cell formation problem with multiple objectives is more difficult than single objective problem. The paper considers multiple objectives; minimize number of intercell movements, minimize intracell workload variation and minimize intercell workload variation. We propose a multiple objective genetic algorithms(MOGA) resolving the mentioned three objectives. The MOGA procedure adopted Pareto optimal solution for selection method for next generation and the concept of Euclidean distance from the ideal and negative ideal solution for fitness test of a individual. As we consider several weights, decision maker will be reflected his consideration by adjusting high weights for important objective. A numerical example is given for a comparative analysis with the results of other research.

• 전자공학회논문지
• /
• 제50권4호
• /
• pp.3-8
• /
• 2013

최근 들어 CDMA 기반 시스템에서 고속 데이터 서비스 지원을 위한 수신 성능 개선 방법의 하나로 칩 단위에서 동작하는 선형 최소 평균 제곱 오차(LMMSE: linear minimum mean-squared error) 수신기에 대한 연구가 활발히 진행되고 있다. 본 논문에서는 D-TxAA (dual stream transmit antenna array) 방식을 사용하는 단일 스트림(single stream) HSDPA MIMO 시스템에서 선부호화 계수(precoding weight) 적용이 LMMSE 수신기의 성능에 미치는 영향을 분석하고자 한다. 우선 D-TxAA 기반 $2{\times}2$ 다중입력 다중출력(MIMO: multiple input multiple output) 시스템에서 수신 신호 순간 파워 (instantaneous power)의 총 합이 최대가 되도록 하는 선부호화 계수 결정 방식을 유도할 것이다. 또한 선부호화 계수의 적용 지연 시간과 단말의 이동 속도가 LMMSE 수신기의 성능에 미치는 영향을 분석하고, 컴퓨터 모의실험을 통해 검증하고자 한다.

• 한국경영과학회지
• /
• 제37권1호
• /
• pp.29-43
• /
• 2012

We design a DEA-AR model using multiple regression analysis with new methods which limit weights. When there are multiple input and single output variables, our model can be used, and the weights of input variables use the regression coefficient and coefficient of determination. To verify the effectiveness of the new model, we evaluate the efficiency of the Regional Corporations in Korea. Accordance with statistical analysis, it proved that there is no difference between the efficiency value of the DEA-AR using AHP and our DEA-AR model. Our model can be applied to a lot of research by substituting DEA-AR model relying on AHP in the future.

• 한국멀티미디어학회논문지
• /
• 제20권8호
• /
• pp.1406-1420
• /
• 2017

Artificial neural networks (ANNs) play an important role in the fields of function approximation, prediction, and classification. ANN performance is critically dependent on the input parameters, including the number of neurons in each layer, and the optimal values of weights and biases assigned to each neuron. In this study, we apply the particle swarm optimization method, a popular optimization algorithm for determining the optimal values of weights and biases for every neuron in different layers of the ANN. Several regression models, including general linear regression, Fourier regression, smoothing spline, and polynomial regression, are conducted to evaluate the proposed method's prediction power compared to multiple linear regression (MLR) methods. In addition, residual analysis is conducted to evaluate the optimized ANN accuracy for both training and test datasets. The experimental results demonstrate that the proposed method can effectively determine optimal values for neuron weights and biases, and high accuracy results are obtained for prediction applications. Evaluations of the proposed method reveal that it can be used for prediction and estimation purposes, with a high accuracy ratio, and the designed model provides a reliable technique for optimization. The simulation results show that the optimized ANN exhibits superior performance to MLR for prediction purposes.

• 대한수학회보
• /
• 제48권3호
• /
• pp.445-454
• /
• 2011

In this paper, we first find a raising operator and a lowering operator for multiple Bessel polynomials and then give a differential equation having multiple Bessel polynomials as solutions. Thus the differential equations were found for all multiple orthogonal polynomials that are orthogonal with respect to the same type of classical weights introduced by Aptekarev et al.