This paper addresses the robust ℋ∞ static output feedback (SOF) controller design problem for a class of uncertain fuzzy affine systems that are robust against both the plant parameter perturbations and controller gain variations. More specifically, the purpose is to synthesize a non-fragile piecewise affine SOF controller guaranteeing the stability of the resulting closed-loop fuzzy affine dynamic system with certainℋ∞ performance index. Based on piecewise quadratic Lyapunov functions and applying some convexification procedures, two different approaches are proposed to solve the robust and non-fragile piecewise affine SOF controller synthesis problem. It is shown that the piecewise affine controller gains can be obtained by solving a set of linear matrix inequalities (LMIs). Finally, simulation examples are given to illustrate the effectiveness of the proposed methods.
The stochastic interpretation of Parikh’s game logic should not follow the usual pattern of Kripke models, which in turn are based on the Kleisli morphisms for the Giry monad, rather, a specific and more general approach to probabilistic nondeterminism is required.We outline this approach together with its probabilistic and measure theoretic basis, introducing in a leisurely pace the Giry monad and its Kleisli morphisms together with important techniques for manipulating them. Proof establishing specific techniques are given, and pointers to the extant literature are provided.
After working through this tutorial, the reader should find it easier to follow the original literature in this and related areas, and it should be possible for her or him to appreciate measure theoretic arguments for original work in the areas of Markov transition systems, and stochastic effectivity functions.
Program slicing is an effective technique for analyzing concurrent programs. However, when a conventional closure-based slicing algorithmfor sequential programs is applied to a concurrent interprocedural program, the slice is usually imprecise owing to the intransitivity of interference dependence. Interference dependence arises when a statement uses a variable defined in another statement executed concurrently. In this study, we propose a global dependence analysis approach based on a program reachability graph, and construct a novel dependence graph calledmarking-statement dependence graph (MSDG), in which each vertex is a 2-tuple of program state and statement. In contrast to the conventional program dependence graph where the vertex is a statement, the dependence relation in MSDG is transitive. When traversing MSDG, a precise slice will be obtained. To enhance the slicing efficiency without loss of precision, our slicing algorithm adopts a hybrid strategy. The procedures containing interaction statements between threads are inlined and sliced by the slicing algorithm based on program reachability graphs while allowing other procedures to be sliced as sequential programs. We have implemented our algorithm and three other representative slicing algorithms, and conducted an empirical study on concurrent Java programs. The experimental results show that our algorithm computes more precise slices than the other algorithms. Using partial-order reduction techniques, which are effective for reducing the size of a program reachability graph without loss of precision, our algorithm is optimized, thereby improving its performance to some extent.
Context-aware system is an emerging research area in recent years. Context plays an important role in these systems. In most existing work, context is treated as all relative elements in the environment of an application, and the scope of context is predefined by the developers during the development. However, it is difficult to analyze, specify, and organize everything in the environment accurately and completely; and even when it is possible, the developed applications are difficult to extend or modify as the requests for environment may change over time. In this paper, we focus on activity-oriented context-aware (AOCA) applications where the requests for environment are highly dependent on user activities, and propose a programming framework for developing AOCA applications. In particular, we first present a concept model for describing the notions of activity-oriented context. Next, based on the concept model, we describe the details of the programming framework as well as a development tool. Moreover, we provide a platform to support the runtime of AOCA applications, and demonstrate the advantages of our programming framework through experimental evaluations.
Instagram is a popular photo-sharing social application. It is widely used by tourists to record their journey information such as location, time and interest. Consequently, a huge volume of geo-tagged photos with spatio-temporal information are generated along tourist’s travel trajectories. Such Instagram photo trajectories consist of travel paths, travel density distributions, and traveller behaviors, preferences, and mobility patterns. Mining Instagram photo trajectories is thus very useful for many mobile and location-based social applications, including tour guide and recommender systems. However, we have not found any work that extracts interesting group-like travel trajectories from Instagram photos asynchronously taken by different tourists. Motivated by this, we propose a novel concept:coterie, which reveals representative travel trajectory patterns hidden in Instagram photos taken by users at shared locations and paths. Our work includes the discovery of (1)coteries, (2) closed coteries, and (3) the recommendation of popular travel routes based on closed coteries. For this, we first build a statistically reliable trajectory database from Instagram geo-tagged photos. These trajectories are then clustered by the DBSCAN method to find tourist density. Next, we transform each raw spatio-temporal trajectory into a sequence of clusters. All discriminative closedcoteriesare further identified by a Cluster-Growth algorithm. Finally, distance-aware and conformityaware recommendation strategies are applied onclosed coteriesto recommend popular tour routes. Visualized demos and extensive experimental results demonstrate the effectiveness and efficiency of our methods.
Color descriptors of an image are the most widely used visual features in content-based image retrieval systems. In this study, we present a novel color-based image retrieval framework by integrating color space quantization and feature coding. Although color features have advantages such as robustness and simple extraction, direct processing of the abundant amount of color information in an RGB image is a challenging task. To overcome this problem, a color space clustering quantization algorithm is proposed to obtain the clustering color space (CCS) by clustering the CIE1976L∗a∗b∗ space into 256 distinct colors, which adequately accommodate human visual perception. In addition, a new feature coding method called feature-to-character coding (FCC) is proposed to encode the block-based main color features into character codes. In this method, images are represented by character codes that contribute to efficiently building an inverted index by using color features and by utilizing text-based search engines. Benefiting from its high-efficiency computation, the proposed framework can also be applied to large-scale web image retrieval. The experimental results demonstrate that the proposed system can produce a significant augmentation in performance when compared to blockbased main color image retrieval systems that utilize the traditional HSV(Hue, Saturation, Value) quantization method.
In the context of robotics, configuration space (cspace) is widely used for non-circular robots to engage tasks such as path planning, collision check, and motion planning. In many real-time applications, it is important for a robot to give a quick response to the user’s command. Therefore, a constant bound on planning time per action is severely imposed. However, existing search algorithms used in c-space gain first move lags which vary with the size of the underlying problem. Furthermore, applying real-time search algorithms on c-space maps often causes the robots being trapped within local minima. In order to solve the above mentioned problems, we extend the learning real-time search (LRTS) algorithm to search on a set of c-space generalized Voronoi diagrams (c-space GVDs), helping the robots to incrementally plan a path, to efficiently avoid local minima, and to execute fast movement. In our work, an incremental algorithm is firstly proposed to build and represent the c-space maps in Boolean vectors. Then, the method of detecting grid-based GVDs from the c-space maps is further discussed. Based on the c-space GVDs, details of the LRTS and its implementation considerations are studied. The resulting experiments and analysis show that, using LRTS to search on the c-space GVDs can 1) gain smaller and constant first move lags which is independent of the problem size; 2) gain maximal clearance from obstacles so that collision checks are much reduced; 3) avoid local minima and thus prevent the robot from visually unrealistic scratching.
Target tracking applications of wireless sensor networks (WSNs) may provide a high performance only when a reliable collection of target positions from sensor nodes is ensured. The performance of target tracking in WSNs is affected by transmission delay, failure probability, and nodes energy depletion. These negative factors can be effectively mitigated by decreasing the amount of transmitted data. Thus, the minimization of data transfers from sensor nodes is an important research issue for the development of WSN-based target tracking applications. In this paper, a data suppression approach is proposed for target chasing in WSNs. The aim of the considered target chasing task is to catch a moving target by a mobile sink in the shortest time. According to the introduced approach, a sensor node sends actual target position to the mobile sink only if this information is expected to be useful for minimizing the time in which target will be caught by the sink. The presented method allows sensor nodes to evaluate the usefulness of sensor readings and select those readings that have to be reported to the sink. Experiments were performed in a simulation environment to compare effectiveness of the proposed approach against state-of-the-art methods. Results of the experiments show that the presented suppression method enables a substantial reduction in the amount of transmitted data with no significant negative effect on target chasing time.
Transmission control protocol (TCP) is a reliable transport layer protocol widely used in the Internet over decades. However, the performances of existing TCP congestion control algorithms degrade severely in modern heterogeneous networks with random packet losses, packet reordering and congestion. In this paper, we propose a novel TCP algorithm named TCP-ACC to handle all three challenges mentioned above. It integrates 1) a real-time reorder metric for calculating the probabilities of unnecessary Fast Retransmit (FRetran) and Timeouts (TO), 2) an improved RTT estimation algorithm giving more weights to packets that are sent (as opposed to received) more recently, and 3) an improved congestion control mechanism based on packet loss and reorder rate measurements. Theoretical analysis demonstrates the equilibrium throughput of TCP-ACC is much higher than traditional TCP, while maintaining good fairness with regard to other TCP algorithms in ideal network conditions. Extensive experimental results using both network emulators and real network show that the algorithm achieves significant throughput improvement in heterogeneous networks as compared with other state-of-the-art algorithms.
The increasing availability of folksonomy data makes them vital for user profiling approaches to precisely detect user preferences and better understand user interests, so as to render some personalized recommendation or retrieval results. This paper presents a rigorous probabilistic framework to discover user preference from folksonomy data. Furthermore, we incorporate three models into the framework with the corresponding inference methods, expectation-maximization or Gibbs sampling algorithms. The user preference is expressed through topical conditional distributions. Moreover, to demonstrate the versatility of our framework, a recommendation method is introduced to show the possible usage of our framework and evaluate the applicability of the engaged models. The experimental results show that, with the help of the proposed framework, the user preference can be effectively discovered.
Personal profile information on social media like LinkedIn.com and Facebook.com is at the core of many interesting applications, such as talent recommendation and contextual advertising. However, personal profiles usually lack consistent organization confronted with the large amount of available information. Therefore, it is always a challenge for people to quickly find desired information from them. In this paper, we address the task of personal profile summarization by leveraging both textual information and social connection information in social networks from both unsupervised and supervised learning paradigms. Here, using social connection information is motivated by the intuition that people with similar academic, business or social background (e.g., comajor, co-university, and co-corporation) tend to have similar experiences and should have similar summaries. For unsupervised learning, we propose a collective ranking approach, called SocialRank, to combine textual information in an individual profile and social context information from relevant profiles in generating a personal profile summary. For supervised learning, we propose a collective factor graph model, called CoFG, to summarize personal profiles with local textual attribute functions and social connection factors. Extensive evaluation on a large dataset from LinkedIn.com demonstrates the usefulness of social connection information in personal profile summarization and the effectiveness of our proposed unsupervised and supervised learning approaches.
Modern search engines record user interactions and use them to improve search quality. In particular, user click-through has been successfully used to improve clickthrough rate (CTR), Web search ranking, and query recommendations and suggestions. Although click-through logs can provide implicit feedback of users’ click preferences, deriving accurate absolute relevance judgments is difficult because of the existence of click noises and behavior biases. Previous studies showed that user clicking behaviors are biased toward many aspects such as “position” (user’s attention decreases from top to bottom) and “trust” (Web site reputations will affect user’s judgment). To address these problems, researchers have proposed several behavior models (usually referred to as click models) to describe users? practical browsing behaviors and to obtain an unbiased estimation of result relevance. In this study, we review recent efforts to construct click models for better search ranking and propose a novel convolutional neural network architecture for building click models. Compared to traditional click models, our model not only considers user behavior assumptions as input signals but also uses the content and context information of search engine result pages. In addition, our model uses parameters from traditional click models to restrict the meaning of some outputs in our model’s hidden layer. Experimental results show that the proposed model can achieve considerable improvement over state-of-the-art click models based on the evaluation metric of click perplexity.
In this paper, a new scheme that uses digraph substitution rules to conceal the mechanism or activity required to derive password-images is proposed. In the proposed method, a user is only required to click on one of the pass-image instead of both pass-images shown in each challenge set for three consecutive sets.While this activity is simple enough to reduce login time, the images clicked appear to be random and can only be obtained with complete knowledge of the registered password along with the activity rules. Thus, it becomes impossible for shoulder-surfing attackers to obtain the information about which password images and pass-images are used by the user. Although the attackers may know about the digraph substitution rules used in the proposed method, the scenario information used in each challenge set remains. User study results reveal an average login process of less than half a minute. In addition, the proposed method is resistant to shoulder-surfing attacks.