Nguyen Hoang Viet

Kyung Hee University, Sŏul, Seoul, South Korea

Are you Nguyen Hoang Viet?

Claim your profile

Publications (18)0.74 Total impact

  • Ngo Anh Vien · Nguyen Hoang Viet · SeungGwan Lee · TaeChoong Chung
    [Show abstract] [Hide abstract]
    ABSTRACT: In this paper, we solve the call admission control (CAC) and routing problem in an integrated network that handles several classes of calls of different values and with different resource requirements. The problem of maximizing the average reward (or cost) of admitted calls per unit time is naturally formulated as a semi-Markov Decision Process (SMDP) problem, but is too complex to allow for an exact solution. Thus in this paper, a policy gradient algorithm, together with a decomposition approach, is proposed to find the dynamic (state-dependent) optimal CAC and routing policy among a parameterized policy space. To implement that gradient algorithm, we approximate the gradient of the average reward. Then, we present a simulation-based algorithm to estimate the approximate gradient of the average reward (called GSMDP algorithm), using only a single sample path of the underlying Markov chain for the SMDP of CAC and routing problem. The algorithm enhances performance in terms of convergence speed, rejection probability, robustness to the changing arrival statistics and an overall received average revenue. The experimental simulations will compare our method's performance with other existing methods and show the robustness of our method.
    No preview · Article · Jun 2009 · IEICE Transactions on Communications
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Prediction problems are prevalent in medical domains. For exam- ple, computer-aided diagnosis or prognosis is a key component in a CDSS (Clinical Decision Support System). SVMs, especially SVMs with nonlinear kernels such RBF kernels, have shown supe- rior accuracy in prediction problems. However, they are not favor- ably used by physicians for medical prediction problems because nonlinear SVMs are difficult to visualize, thus it is hard to provide intuitive interpretation of prediction results to physicians. Nomo- gram was proposed to visualize SVM classification models. How- ever, it cannot visualize nonlinear SVM models. Localized RBF (LRBF) kernel was proposed which shows comparable accuracy as the RBF kernel while the LRBF kernel is easier to interpret since it can be linearly decomposed. This paper presents a new tool named VRIFA, which integrates the nomogram and LRBF kernel to pro- vide users with an interactive visualization of nonlinear SVM mod- els. VRIFA graphically exposes the internal structure of nonlinear SVM models showing the effect of each feature, the magnitude of the effect, and the change at the prediction output. VRIFA also per- forms nomogram-based feature selection while training a model in order to remove noise or redundant features and improve the pre- diction accuracy. The tool has been used by biomedical researchers for computer-aided diagnosis and risk factor analysis for diseases. VRIFA is accessible at http://dm.postech.ac.kr/vrifa .
    Full-text · Conference Paper · Jan 2009
  • [Show abstract] [Hide abstract]
    ABSTRACT: Recently, Support Vector Machines (SVMs) have been applied very effectively in learning ranking functions (or preference functions).They intend to learn ranking functions with the principles of the large margin and the kernel trick. However, the output of a ranking function is a score function which is not a calibrated posterior probability to enable post-processing. One approach to deal with this problem is to apply a generalized linear model with a link function and solve it by calculating the maximum likelihood estimate. But, if the link function is nonlinear, maximizing the likelihood will face with difficulties. Instead, we propose a new approach which train an SVM for a ranking function, then map the SVM outputs into a probabilistic sigmoid function whose parameters are trained by using cross-validation. This method will be tested on three data-mining datasets and compared to the results obtained by standard SVMs.
    No preview · Conference Paper · Jan 2009
  • Nguyen Hoang Viet · Ngo Anh Vien · TaeChoong Chung
    [Show abstract] [Hide abstract]
    ABSTRACT: This paper presents a policy gradient semi-Markov decision process (SMDP) algorithm for call admission control and routing functions in an integrated network. These systems must handle several classes of calls of different value and with different resource requirements. The problem of maximizing the average reward (or cost) of admitted calls per unit time is naturally formulated as a SMDP problem, but is too complex to allow for an exact solution. Thus, a policy gradient algorithm is proposed to find the optimal call admission control and routing policy among a parameterized randomized policy space. To implement that gradient algorithm, we approximate the gradient of the average reward. Then, we present a simulation-based algorithm to estimate the approximate average gradient of the average reward (GSMDP), using only single sample path of the underlying Markov chain for the SMDP of call admission control and routing problem. The experimental simulations will compare its performance with other methods show the robustness of our algorithm.
    No preview · Conference Paper · May 2008
  • Nguyen Hoang Viet · Ngo Anh Vien · SeungGwan Lee · TaeChoong Chung
    [Show abstract] [Hide abstract]
    ABSTRACT: In this paper, a hybrid sensor network consisting of static and mobile sensors is considered, where static sensors are used to detect events, and mobiles sensors can move to event locations to conduct more advanced analysis. This dispatching problem can be converted to maximum weighted maximum matching problem with the objective of minimizing the total energy consumption to move sensor or maximizing the average remaining energy of sensor after movement. We present a distributed algorithm for sensor dispatch using message passing rule. Simulation results show that the algorithm can solve sensor dispatch problem in optimal distributed way.
    No preview · Conference Paper · Apr 2008
  • Nguyen Hoang Viet · Ngo Anh Vien · SeungGwan Lee · TaeChoong Chung
    [Show abstract] [Hide abstract]
    ABSTRACT: The task of planning trajectories for a mobile robot has received considerable attention in the research literature. The problem involves computing a collision-free path between a start point and a target point in environment of known obstacles. In this paper, we study an obstacle avoidance path planning problem using multi ant colony system, in which several colonies of ants cooperate in finding good solution by exchanging good information. In the simulation, we experimentally investigate the behaviour of multi colony ant algorithm with different kinds of information among the colonies. At last we will compare the behaviour of different number of colonies with a multi start single colony ant algorithm to show the good improvement.
    No preview · Conference Paper · Mar 2008
  • Ngo Anh Vien · Nguyen Hoang Viet · Hyun Kim · SeungGwan Lee · TaeChoong Chung
    [Show abstract] [Hide abstract]
    ABSTRACT: This paper introduces ant colony system (ACS), a distributed algorithm that is applied to the Stable Marriage Problem (SM). The stable marriage problem is an extensively-studied combinatorial problem with many practical applications. It is well known that at least one stable matching exists for every stable marriage instance. However, the classical Gale-Shapley [2] algorithm produces a marriage that greatly favors the men at the expense of the women, or vice versa. In our proposed ACS, a set of cooperating agents called ants cooperate to find stable matchings such as stable matching with man-optimal, woman-optimal, egalitarian stable matching, sex-fair stable matching. So this ACS is a novel method to solve Stable Marriage Problem. Our simulation results show the effectiveness of the proposed ACS.
    No preview · Chapter · Aug 2007
  • [Show abstract] [Hide abstract]
    ABSTRACT: In this paper, the Q-Learning based univector field method is proposed for mobile robot to accomplish the obstacle avoidance and the robot orientation at the target position. Univector field method guarantees the desired posture of the robot at the target position. But it does not navigate the robot to avoid obstacles. To solve this problem, modified univector field is used and trained by Q-learning. When the robot following the field to get the desired posture collides with obstacles, univector fields at collision positions are modified according to the reinforcement of Q-learning algorithm. With this proposed navigation method, robot navigation task in a dynamically changing environment becomes easier by using double action Q-learning [8] to train univector field instead of ordinary Q-learning. Computer simulations and experimental results are carried out for an obstacle avoidance mobile robot to demonstrate the effectiveness of the proposed scheme.
    No preview · Chapter · Aug 2007
  • Nguyen Hoang Viet · Michał Kleiber
    [Show abstract] [Hide abstract]
    ABSTRACT: In this paper systems of linear equations Ax = b, where both A and b contain uncertain factors in terms of fuzziness are investigated. The classical solutions being vectors of fuzzy numbers are considered. The complex problem of finding the exact classical solutions is replaced by a corresponding optimization task with the cost function based on the Hausdorff metric. This cost function is next minimized with use of genetic algorithms. A number of numerical experiments are provided in order to verify the given approach. The results and some conclusions are also included.
    No preview · Chapter · Jul 2007
  • Ngo Anh Vien · Nguyen Hoang Viet · SeungGwan Lee · TaeChoong Chung
    [Show abstract] [Hide abstract]
    ABSTRACT: In this paper, we consider reinforcement learning in systems with unknown environment where the agent must trade off efficiently between: exploration(long-term optimization) and exploitation (short-term optimization). ε− greedy algorithm is a method using near-greedy action selection rule. It behaves greedily (exploitation) most of the time, but every once in a while, say with small probability ε (exploration), instead select an action at random. Many works already proved that random exploration drives the agent towards poorly modeled states. Therefore, this study evaluates the role of heuristic based exploration in reinforcement learning. We proposed three methods: neighborhood search based exploration, simulated annealing based exploration, and tabu search based exploration. All techniques follow the same rule ”Explore the most unvisited state”. In the simulation, these techniques are evaluated and compared on a discrete reinforcement learning task (robot navigation).
    No preview · Conference Paper · Jun 2007
  • Ngo Anh Vien · Nguyen Hoang Viet · SeungGwan Lee · TaeChoong Chung
    [Show abstract] [Hide abstract]
    ABSTRACT: Path planning is an important task in mobile robot control. When the robot must move rapidly from any arbitrary start positions to any target positions in environment, a proper path must avoid both static obstacles and moving obstacles of arbitrary shape. In this paper, an obstacle avoidance path planning approach for mobile robots is proposed by using Ant-Q algorithm. Ant-Q is an algorithm in the family of ant colony based methods that are distributed algorithms for combinatorial optimization problems based on the metaphor of ant colonies. In the simulation, we experimentally investigate the sensitivity of the Ant-Q algorithm to its three methods of delayed reinforcement updating and we compare it with the results obtained by other heuristic approaches based on genetic algorithm or traditional ant colony system. At last, we will show very good results obtained by applying Ant-Q to bigger problem: Ant-Q find very good path at higher convergence rate.
    No preview · Conference Paper · Jun 2007
  • Nguyen Hoang Viet · Michal Kleiber

    No preview · Conference Paper · Jan 2007
  • Nguyen Hoang Viet · Michal Kleiber
    [Show abstract] [Hide abstract]
    ABSTRACT: The problem of solving systems of interval linear equations with use of AI based approaches is studied in this paper. First, this problem is viewed in terms of an optimization task. A cost function with interval variables is defined. Next, for a given system of equations, instead of the exact algebraic solution its approximation is determined by minimizing the cost function. This is done by use of two different approaches: the NN based approach and the GA based one. A number of numerical evaluations are provided in order to verify the proposed techniques. The results are compared, discussed and some final conclusions are drawn.
    No preview · Conference Paper · Jan 2006
  • Nguyen Hoang Viet · Michal Kleiber
    [Show abstract] [Hide abstract]
    ABSTRACT: A new approach for approximating the algebraic solution of systems of interval linear equations, denoted here as SILE, is proposed. The original SILE problem is first considered in terms of an optimization problem. The exact arithmetic operations over interval numbers which appear in the cost function of this optimization problem are replaced by those performed with the use of some simple sigmoidal feed-forward neural networks. This modified cost function is then minimized using one of the gradient based algorithms. A number of numerical evaluations are provided in order to verify the proposed approach. The results are discussed and some final remarks are included.
    No preview · Article · Jan 2005
  • Nguyen Hoang Viet · Michał Kleiber
    [Show abstract] [Hide abstract]
    ABSTRACT: A new approach to approximate the algebraic solution of systems of interval linear equations (SILE) is proposed in this paper. The original SILE problem is first transformed into an optimization problem, which is in turn solved with use of artificial neural networks and gradient-based optimization techniques.
    No preview · Chapter · Dec 2004
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: We have designed and constructed a scalable and reconfigurable optical IP network system for use in residential premises. The network system consists of an optical fiber bus, bus interfaces attached to the bus, and a gateway. In the newly constructed network, 8 optical wavelengths in the course WDM standard were allocated to Gigabit optical transceivers and optical add/drop modules, respectively. The performance examination clarified that throughput of around 90 Mbps was obtained even when 8 personal computers belonging to the different 8 sub networks sent packets simultaneously. The scalability was also confirmed by experiment.
    Preview · Conference Paper · Feb 2004
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: A novel optical network architecture to integrate layer 2 and 3 routings by optical/electrical hybrid switching is proposed. One node router and 8 edge routers were prepared and connected with single-mode fibers so as to form an optical ring, and the transmission experiments were conducted. It was successfully confirmed in the performance examination that layer 2 and 3 transmission could be integrated through the network, by using 8 wavelengths in 1550 nm band for the layer 2 transmission and 1310 nm wavelength for the layer 3 transmission.
    Preview · Conference Paper · Oct 2003
  • Source
    Nguyen Hoang Viet · Jacek Mańdziuk
    [Show abstract] [Hide abstract]
    ABSTRACT: In this work several approaches to prediction of natural gas consumption with neural and fuzzy neural systems for a certain region of Poland are analyzed and tested. Prediction strategies tested in the paper include: single neural net module approach, combination of three neural modules, temperature clusterization based method, and application of fuzzy neural networks. The results indicate the superiority of temperature clusterization based method over modular and fuzzy neural approaches. One of the interesting issues observed in the paper is relatively good performance of the tested methods in the case of a long-term (four week) prediction compared to mid-term (one week) prediction. Generally, the results are significantly better than those obtained by statistical methods currently used in the gas company under consideration.
    Full-text · Conference Paper · Oct 2003