• 대한음성학회지:말소리
• /
• 제45호
• /
• pp.117-126
• /
• 2003

Decision tree-based state tying has been proposed in recent years as the most popular approach for clustering the states of context-dependent hidden Markov model-based speech recognition. The aims of state tying is to reduce the number of free parameters and predict state probability distributions of unseen models. But, when doing state tying, the size of a decision tree is very important for word independent recognition. In this paper, we try to construct optimized decision tree based on the average of feature vectors in state pool and the number of seen modes. We observed that the proposed optimal decision tree is effective in predicting the state probability distribution of unseen models.

• 한국통신학회논문지
• /
• 제32권10A호
• /
• pp.1004-1014
• /
• 2007

In sensor networks, it is crucial to reliably and energy-efficiently deliver sensed information from each source to a sink node. Specifically, in mobile sink (user) applications, due to the sink mobility, a stationary dissemination path may no longer be effective. The path will have to be continuously reconfigured according to the current location of the sink. Moreover, the dynamic optimal path from each source to the sink is required in order to reduce end-to-end delay and additional energy wastage. In this paper, an Adaptive Reversal Optimal path Tree (AROT) protocol is proposed. Information delivery from each source to a mobile sink can be easily achieved along the AROT without additional control overhead, because the AROT proactively performs adaptive sink mobility management. In addition, the dynamic path is optimal in terms of hop counts and the AROT can maintain a robust tree structure by quickly recovering the partitioned tree with minimum packet transmission. Finally, the simulation results demonstrate that the AROT is a considerably energy-efficient and robust protocol.

• 한국정보통신학회논문지
• /
• 제8권5호
• /
• pp.1089-1095
• /
• 2004

주어진 네트워크에서 최적의 스타이너 트리를 구하는 문제는 NP-hard이며, 최적에 가까운 스타이너 트리를 구하기 위하여 유전자 알고리즘을 이용한다. 본 논문에서는 이 문제를 해결하기 위하여 유전자 알고리즘에서 염색체를 기존의 이진스트링 대신 부동소수점으로 표현하였다. 먼저 주어진 네트워크에 Prim의 알고리즘을 적용하여 스패닝 트리를 구하고, 부동소수점 표현을 갖는 유전자 알고리즘을 사용하여 새로운 스타이너 점을 트리에 추가하는 과정을 반복함으로써 최적에 가까운 스타이너 트리를 구했다 이 방법을 사용하면 이진스트링을 사용하는 기존의 방법에 비해서 트리가 보다 빠르고 정확하게 최적에 가까운 스타이너 트리에 접근했다.

• 한국경영과학회지
• /
• 제40권2호
• /
• pp.19-29
• /
• 2015

The purpose of this study is to suggest the grid search method for selecting an optimal decision tree model. It chooses optimal values for the maximum depth of tree and the minimum number of observations that must exist in a node in order for a split to be attempted. Therefore, the grid search method guarantees building a decision tree model that shows more precise and stable classifying performance. Through empirical analysis using data of job satisfaction of workplace reserve force commanders, we show that the grid search method helps us generate an optimal decision tree model that gives us hints for the improvement direction of labor conditions of Korean workplace reserve force commanders.

• East Asian mathematical journal
• /
• 제38권3호
• /
• pp.277-292
• /
• 2022

This study investigates the convergence of the optimal consumption and investment policies in a binomial-tree model to those in the continuous-time model of Merton (1969). We provide the convergence in explicit form and show that the convergence rate is of order ∆t, which is the length of time between consecutive time points. We also show by numerical solutions with realistic parameter values that the optimal policies in the binomial-tree model do not differ significantly from those in the continuous-time model for long-term portfolio management with a horizon over 30 years if rebalancing is done every 6 months.

• 한국정보시스템학회:학술대회논문집
• /
• 한국정보시스템학회 2002년도 추계학술대회
• /
• pp.289-312
• /
• 2002

This study suggests the Link weight analysis approach to choose input variables and an integrated model to make more accurate bankruptcy prediction model. the Link weight analysis approach is a method to choose input variables to analyze each input node's link weight which is the absolute value of link weight between an input nodes and a hidden layer. There are the weak-linked neurons elimination method, the strong-linked neurons selection method in the link weight analysis approach. The Integrated Model is a combined type adapting Bagging method that uses the average value of the four models, the optimal weak-linked-neurons elimination method, optimal strong-linked neurons selection method, decision-making tree model, and MDA. As a result, the methods suggested in this study - the optimal strong-linked neurons selection method, the optimal weak-linked neurons elimination method, and the integrated model - show much higher accuracy than MDA and decision making tree model. Especially the integrated model shows much higher accuracy than MDA and decision making tree model and shows slightly higher accuracy than the optimal weak-linked neurons elimination method and the optimal strong-linked neurons selection method.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2017년도 춘계학술발표대회
• /
• pp.677-679
• /
• 2017

Data convergecast is an indispensable task for any WSN applications. Typically, scheduling in the WSN consists of two phases: tree construction and scheduling. The optimal tree structure and scheduling for the network is proven NP-hard. This paper focuses on the delay optimality while constructing the data convergecast tree. The algorithm can take any tree as the input, and by performing the switches (i.e. a node changes its parent), the expected aggregation delay is potentially reduced. Note that while constructing the tree, only the in-tree collisions between the child nodes sending data to their common parent is considered.

• Journal of applied mathematics & informatics
• /
• 제5권2호
• /
• pp.427-432
• /
• 1998

When an undirected tree T and a vertex ${\gamma}$ in the tree are given the problem to transform T into a rooted tree with ${\gamma}$ as its root is considered. Using Euler tour and prefix sum an optimal algorithm has been developed [2,3]. We will present another parallel algorithm which is optimal also on EREW PRAM. Our approach resuces the given tree step by step by pruning and pointer jumping. That is the tree structure is retained during algorithm processing such that than other tree computations can be carried out in parallel.

• 한국환경복원기술학회지
• /
• 제22권2호
• /
• pp.1-12
• /
• 2019

Open space planners and designers should consider scientific and quantified functions of trees when they have to locate where to plant the tree. However, until now, most planners and designers could not consider them because of lack of tool for considering scientific and quantitative tree functions. This study introduces a tree location supporting tool which focuses on the multi-objective including scientific function using ACO (Ant colony optimization). We choose shading effect (scientific function), shielding, and openness as objectives for test application. The results show that when the user give a high weight to a particular objective, they can obtain the optimal results with high value of that objective. When we allocate higher weight for the shading effect, the tree plans provide larger shadow value. Even when compared with current tree plan, the study result has a larger shading effect plan. This result will reduce incident radiation to the ground and make thermal friendly open space in the summer. If planners and designers utilize this tool and control the objectives, they would get diverse optimal tree plans and it will allow them to make use of the many environmental benefits from trees.

• 한국컴퓨터정보학회논문지
• /
• 제21권5호
• /
• pp.79-89
• /
• 2016

In this paper, we proposed the object classification method using genetic and dynamic random forest consisting of optimal combination of unit tree. The random forest can ensure good generalization performance in combination of large amount of trees by assigning the randomization to the training samples and feature selection, etc. allocated to the decision tree as an ensemble classification model which combines with the unit decision tree based on the bagging. However, the random forest is composed of unit trees randomly, so it can show the excellent classification performance only when the sufficient amounts of trees are combined. There is no quantitative measurement method for the number of trees, and there is no choice but to repeat random tree structure continuously. The proposed algorithm is composed of random forest with a combination of optimal tree while maintaining the generalization performance of random forest. To achieve this, the problem of improving the classification performance was assigned to the optimization problem which found the optimal tree combination. For this end, the genetic algorithm methodology was applied. As a result of experiment, we had found out that the proposed algorithm could improve about 3~5% of classification performance in specific cases like common database and self infrared database compare with the existing random forest. In addition, we had shown that the optimal tree combination was decided at 55~60% level from the maximum trees.