Volume 9, Number 2, 2012
We propose an approach for dependence tree structure learning via copula. A nonparametric algorithm for copula estimation is presented. Then a Chow-Liu like method based on dependence measure via copula is proposed to estimate maximum spanning bivariate copula associated with bivariate dependence relations. The main advantage of the approach is that learning with empirical copula focuses on dependence relations among random variables, without the need to know the properties of individual variables as well as without the requirement to specify parametric family of entire underlying distribution for individual variables. Experiments on two real-application data sets show the effectiveness of the proposed method.
In modern computer games, bots-intelligent realistic agents play a prominent role in the popularity of a game in the market. Typically, bots are modeled using finite-state machine and then programmed via simple conditional statements which are hard-coded in bots logic. Since these bots have become quite predictable to an experienced games0 player, a player might lose interest in the game. We propose the use of a game theoretic based learning rule called fictitious play for improving behavior of these computer game bots which will make them less predictable and hence, more a enjoyable game.
This paper proposes an adaptive neural network control method for a class of perturbed strict-feedback nonlinear systems with unknown time delays. Radial basis function neural networks are used to approximate unknown intermediate control signals. By constructing appropriate Lyapunov-Krasovskii functionals, the unknown time delay terms have been compensated. Dynamic surface control technique is used to overcome the problem of explosion of complexity in backstepping design procedure. In addition, the semiglobal uniform ultimate boundedness of all the signals in the closed-loop system is proved. A main advantage of the proposed controller is that both problems of curse of dimensionality and explosion of complexity are avoided simultaneously. Finally, simulation results are presented to demonstrate the effectiveness of the approach.
Resource reconstruction algorithms are studied in this paper to solve the problem of resource on-demand allocation and improve the efficiency of resource utilization in virtual computing resource pool. Based on the idea of resource virtualization and the analysis of the resource status transition, the resource allocation process and the necessity of resource reconstruction are presented. Resource reconstruction algorithms are designed to determine the resource reconstruction types, and it is shown that they can achieve the goal of resource on-demand allocation through three methodologies: resource combination, resource split, and resource random adjustment. The effects that the resource users have on the resource reconstruction results, the deviation between resources and requirements, and the uniformity of resource distribution are studied by three experiments. The experiments show that resource reconstruction has a close relationship with resource requirements, but it is not the same with current distribution of resources. The algorithms can complete the resource adjustment with a lower cost and form the logic resources to match the demands of resource users easily.
We present a method to improve the execution time used to build the roadmap in probabilistic roadmap planners. Our method intelligently deactivates some of the configurations during the learning phase and allows the planner to concentrate on those configurations that are most likely going to be useful when building the roadmap. The method can be used with many of the existing sampling algorithms. We ran tests with four simulated robot problems typical in robotics literature. The sampling methods applied were purely random, using Halton numbers, Gaussian distribution, and bridge test technique. In our tests, the deactivation method clearly improved the execution times. Compared with pure random selections, the deactivation method also significantly decreased the size of the roadmap, which is a useful property to simplify roadmap planning tasks.
This paper presents a novel design method for discrete-time repetitive control systems (RCS) based on two-dimensional (2D) discrete-time model. Firstly, the 2D model of an RCS is established by considering both the control action and the learning action in RCS. Then, through constructing a 2D state feedback controller, the design problem of the RCS is converted to the design problem of a 2D system. Then, using 2D system theory and linear matrix inequality (LMI) method, stability criterion is derived for the system without and with uncertainties, respectively. Parameters of the system can be determined by solving the LMI of the stability criterion. Finally, numerical simulations validate the effectiveness of the proposed method.
A Bloom filter is a space-efficient data structure used for concisely representing a set as well as membership queries at the expense of introducing false positive. In this paper, we propose the L-priorities Bloom filter (LPBF) as a new member of the Bloom filter (BF) family, it uses a limited multidimensional bit space matrix to replace the bit vector of standard bloom filters in order to support different priorities for the elements of a set. We demonstrate the time and space complexity, especially the false positive rate of LPBF. Furthermore, we also present a detailed practical evaluation of the false positive rate achieved by LPBF. The results show that LPBF performs better than standard BFs with respect to false positive rate.
The input time delay is always existent in the practical systems. Analysis of the delay phenomenon in a continuous-time domain is sophisticated. It is appropriate to obtain its corresponding discrete-time model for implementation via digital computer. This paper proposes a new discretization method for calculating a sampled-data representation of nonlinear time-delayed non-affine systems. The proposed scheme provides a finite-dimensional representation for nonlinear systems with non-affine time-delayed input enabling existing nonlinear controller design techniques to be applied to them. The performance of the proposed discretization procedure is evaluated by using a nonlinear system with non-affine time-delayed input. For this nonlinear system, various time delay values are considered.
This paper is mainly concerned with the model predictive control (MPC) of networked control systems (NCSs) with uncertain time delay and data packets disorder. The network-induced time delay is described as bounded and arbitrary process. For the usual state feedback controller, by considering all the possibilities of delays, an augmented state space model of the closed-loop system, which characterizes all the delay cases, is obtained. The stability conditions are given according to the Lyapunov method based on this augmented model. The stability property is inherited in MPC which explicitly considers the physical constraints. A numerical example is given to demonstrate the effectiveness of the proposed MPC.
This paper deals with the problem of tracking control for a class of high order nonlinear systems with input delay. The unknown continuous functions of the system are estimated by fuzzy logic systems (FLS). A state conversion method is introduced to eliminate the delayed input item. By means of the backstepping algorithm, the property of semi-globally uniformly ultimately bounded (SGUUB) of the closed-loop system is achieved. The stability of the closed-loop system is proved according to Lyapunov second theorem on stability. The tracking error is proved to be bounded which ultimately converges to an adequately small compact set. Finally, a computer simulation example of high order nonlinear systems is presented, which illustrates the effectiveness of the control scheme.
This paper presents a modification in pulse width modulation (PWM) scheme with unequal shoot-through distribution for the Z-source inverter (ZSI) which can minimize ripples in the current through the Z-source inductors as well as the voltage across the Z-source capacitors. For the same system parameters, the proposed control technique provides better voltage boost across the Z-source capacitor, DC-link, and also the AC output voltage than the traditional PWM. The ripples in the Z-network elements are found to be reduced by 75% in the proposed modulation scheme with optimum harmonic profile in the AC output. Since the Z-network requirement will be based on the ripple profile of the elements, the Z-network requirements can be greatly reduced. The effectiveness of the proposed modulation scheme has been simulated in Matlab/Simulink software and the results are validated by the experiment in the laboratory.
In this paper, global input-to-state stability (ISS) for discrete-time piecewise affine systems with time-delay are considered. Piecewise quadratic ISS-Lyapunov functions are adopted. Both Lyapunov-Razumikhin and Lyapunov-Krasovskii methods are used. The theorems of Lyapunov-Razumikhin type and Lyapunov-Krasovskii type for piecewise affine systems with time-delay are shown, respectively.
Numerous models have been proposed to reduce the classification error of Nave Bayes by weakening its attribute independence assumption and some have demonstrated remarkable error performance. Considering that ensemble learning is an effective method of reducing the classification error of the classifier, this paper proposes a double-layer Bayesian classifier ensembles (DLBCE) algorithm based on frequent itemsets. DLBCE constructs a double-layer Bayesian classifier (DLBC) for each frequent itemset the new instance contained and finally ensembles all the classifiers by assigning different weight to different classifier according to the conditional mutual information. The experimental results show that the proposed algorithm outperforms other outstanding algorithms.
Wireless cooperative communications require appropriate power allocation (PA) between the source and relay nodes. In selfish cooperative communication networks, two partner user nodes could help relaying information for each other, but each user node has the incentive to consume his power solely to decrease its own symbol error rate (SER) at the receiver. In this paper, we propose a fair and efficient PA scheme for the decode-and-forward cooperation protocol in selfish cooperative relay networks. We formulate this PA problem as a two-user cooperative bargaining game, and use Nash bargaining solution (NBS) to achieve a win-win strategy for both partner users. Simulation results indicate that the NBS is fair in that the degree of cooperation of a user only depends on how much contribution its partner can make to decrease its SER at the receiver, and efficient in the sense that the SER performance of both users could be improved through the game.