Two Types of Haar Wavelet Neural Networks for Nonlinear System Identification

Since wavelet transform uses the multi-scale (or multi-resolution) techniques for time series, wavelet transform is suitable for modeling complex signals. Haar wavelet transform is the most commonly used and the simplest one. The Haar wavelet neural network (HWNN) applies the Harr wavelet transform as active functions. It is easy for HWNN to model a nonlinear system at multiple time scales and sudden transitions. In this paper, two types of HWNN, feedforward and recurrent wavelet neural networks, are used to model discrete-time nonlinear systems, which are in the forms of the NARMAX model and state-space model. We first propose an optimal method to determine the structure of HWNN. Then two stable learning algorithms are given for the shifting and broadening coefficients of the wavelet functions. The stability of the identification procedures is proven.     

A scalable multi-attribute hybrid overlay for range queries on the cloud

Cloud computing has become a promising paradigm as next generation computing model, by providing computation, software, data access, and storage services that do not need to know the location of physical resources interconnected across the globe providing such services. In such an environment, important issues as information sharing and resource/service discovery arise. In order to overcome critical limitations in centralized approaches for information sharing and resource/service discovery, this paper proposes a framework of a scalable multi-attribute hybrid overlay featured with decentralized information sharing, flexible resource/service discovery, fault tolerance and load balancing. Additionally, the proposed hybrid overlay integrates a structured P2P system with an unstructured one to support complex queries. Mechanisms such as load balancing and fault tolerance implemented in our proposed system to improve the overall system performance are also discussed. Experimental results show that the performance of the proposed approach is feasible and stable, as the proposed hybrid overlay improves system performance by reducing the number of routing hops and balancing the load by migrating requests.     

Exploring latent browsing graph for question answering recommendation

In this paper, we develop a framework of Question Answering Pages (referred to as QA pages) recommendation. Our proposed framework consists of the two modules: the off-line module to determine the importance of QA pages and the on-line module for on-line QA page recommendation. In the off-line module, we claim that the importance of QA pages could be discovered from user click streams. If the QA pages are of higher importance, many users will click and spend their time on these QA pages. Moreover, the relevant relationships among QA pages are captured by the browsing behavior on these QA pages. As such, we exploit user click streams to model the browsing behavior among QA pages as QA browsing graph structures. The importance of QA pages is derived from our proposed QA browsing graph structures. However, we observe that the QA browsing graph is sparse and that most of the QA pages do not link to other QA pages. This is referred to as a sparsity problem. To overcome this problem, we utilize the latent browsing relations among QA pages to build a QA Latent Browsing Graph. In light of QA latent browsing graph, the importance score of QA pages (referred to as Latent Browsing Rank) and the relevance score of QA pages (referred to as Latent Browsing Recommendation Rank) are proposed. These scores demonstrate the use of a QA latent browsing graph not only to determine the importance of QA pages but also to recommend QA pages. We conducted extensive empirical experiments on Yahoo! Asia Knowledge Plus to evaluate our proposed framework.     

Input‐to‐state stability of min‐max MPC scheme for nonlinear time‐varying delay systems

This paper studies the robustness problem of the min–max model predictive control (MPC) scheme for constrained nonlinear time‐varying delay systems subject to bounded disturbances. The notion of the input‐to‐state stability (ISS) of nonlinear time‐delay systems is introduced. Then by using the Lyapunov–Krasovskii method, a delay‐dependent sufficient condition is derived to guarantee input‐to‐state practical stability (ISpS) of the closed‐loop system by way of nonlinear matrix inequalities (NLMI). In order to lessen the online computational demand, the non‐convex min‐max optimization problem is then converted to a minimization problem with linear matrix inequality (LMI) constraints and a suboptimal MPC algorithm is provided. Finally, an example of a truck‐trailer is used to illustrate the effectiveness of the proposed results. Copyright © 2010 John Wiley and Sons Asia Pte Ltd and Chinese Automatic Control Society     

Optimal Path Problems with Second-Order Stochastic Dominance Constraints

This paper studies optimal path problems integrated with the concept of second order stochastic dominance. These problems arise from applications where travelers are concerned with the trade off between the risks associated with random travel time and other travel costs. Risk-averse behavior is embedded by requiring the random travel times on the optimal paths to stochastically dominate that on a benchmark path in the second order. A general linear operating cost is introduced to combine link- and path-based costs. The latter, which is the focus of the paper, is employed to address schedule costs pertinent to late and early arrival. An equivalent integer program to the problem is constructed by transforming the stochastic dominance constraint into a finite number of linear constraints. The problem is solved using both off-the-shelf solvers and specialized algorithms based on dynamic programming (DP). Although neither approach ensures satisfactory performance for general large-scale problems, the numerical experiments indicate that the DP-based approach provides a computationally feasible option to solve medium-size instances (networks with several thousand links) when correlations among random link travel times can be ignored.     

On the network effect in Web 2.0 applications

In recent years, the proliferation of the world wide web has lead to an increase in a number of applications such as search, social networks and auctions, whose success depends critically upon the number of users of that service. Two examples of such applications are internet auctions and social networks. One of the characteristics of online auctions is that a successful implementation requires a high volume of buyers and sellers at its website. Consequently, auction sites which have a high volume of traffic have an advantage over those in which the volume is limited. This results in even greater polarization of buyers and sellers towards a particular site. The same is true for social networks in which greater use of a given social network increases the use from other participants on the network. This is often referred to as the “network effect” in a variety of interaction-centric applications in networks. While this effect has qualitatively been known to increase the value of the overall network, its effect has never been modeled or studied rigorously. In this paper, we construct a Markov model to analyze the network effect in the case of two important classes of web applications. These correspond to auctions and social networks. We show that the network effect is very powerful and can result in a situation in which an auction or a social networking site can quickly overwhelm its competing sites. Thus, the results of this paper show the tremendous power of the network effect for Web 2.0 applications.     

RBF neural networks-based robust adaptive tracking control for switched uncertain nonlinear systems

In this paper, a robust adaptive H∞ control scheme is presented for a class of switched uncertain nonlinear systems. Radical basis function neural networks (RBF NNs) are employed to approximate unknown nonlinear functions and uncertain terms. A robust H∞ controller is designed to enhance robustness due to the existence of the compound disturbance which consists of approximation errors of the neural networks and external disturbance. Adaptive neural updated laws and switching signals are deducted from multiple Lyapunov function approach. It is proved that with the proposed control scheme, the resulting closed-loop switched system is robustly stable and uniformly ultimately bounded (UUB) such that good capabilities of tracking performance is attained and H∞ tracking error performance index is achieved. A practical example shows the effectiveness of the proposed control scheme.     

Unified uncertainty analysis by the mean value first order saddlepoint approximation

Uncertainties exist in products or systems widely. In general, uncertainties are classified as epistemic uncertainty or aleatory uncertainty. This paper proposes a unified uncertainty analysis (UUA) method based on the mean value first order saddlepoint approximation (MVFOSPA), denoted as MVFOSPA-UUA, to estimate the systems probabilities of failure considering both epistemic and aleatory uncertainties simultaneously. In this method, the input parameters with epistemic uncertainty are modeled using interval variables while input parameters with aleatory uncertainty are modeled using probability distribution or random variables. In order to calculate the lower and upper bounds of system probabilities of failure, both the best case and the worst case scenarios of the system performance function need to be considered, and the proposed MVFOSPA-UUA method can handle these two cases easily. The proposed method is demonstrated to be more efficient, robust and in some situations more accurate than the existing methods such as uncertainty analysis based on the first order reliability method. The proposed method is demonstrated using several examples.     

A semi-Lagrangian level set method for structural optimization

In level set based structural optimization, semi-Lagrange method has an advantage to allow for a large time step without the limitation of Courant–Friedrichs–Lewy (CFL) condition for numerical stability. In this paper, a line search algorithm and a sensitivity modulation scheme are introduced for the semi-Lagrange method. The line search attempts to adaptively determine an appropriate time step in each iteration of optimization. With consideration of some practical characteristics of the topology optimization process, incorporating the line search into semi-Lagrange optimization method can yield fewer design iterations and thus improve the overall computational efficiency. The sensitivity modulation is inspired from the conjugate gradient method in finite-dimensions, and provides an alternative to the standard steepest descent search in level set based optimization. Two benchmark examples are presented to compare the sensitivity modulation and the steepest descent techniques with and without the line search respectively.