Computational intelligence in management of ATM networks

Designing effective control strategies for asynchronous transfer mode (ATM) networks is known to be difficult because of the complexity of the structure of networks, nature of the services supported, and variety of dynamic parameters involved. Additionally, the uncertainties involved in identification of the network parameters cause analytical modeling of ATM networks to be almost impossible. This renders the application of classical control system design methods (which rely on the availability of these models) to the problem even harder. Consequently, a number of researchers are looking at alternative non-analytical control system design and modeling techniques that have the ability to cope with these difficulties to devise effective, robust ATM network management schemes. Those schemes employ artificial neural networks, fuzzy systems and design methods based on evolutionary computation. In this survey, the current state of ATM network management research employing these techniques as reported in the technical literature is summarized. The salient features of the methods employed are reviewed.     

Joint design of clustering and in-cluster data route for heterogeneous wireless sensor networks

A heterogeneous wireless sensor network comprises a number of inexpensive energy constrained wireless sensor nodes which collect data from the sensing environment and transmit them toward the improved cluster head in a coordinated way. Employing clustering techniques in such networks can achieve balanced energy consumption of member nodes and prolong the network lifetimes. In classical clustering techniques, clustering and in-cluster data routes are usually separated into independent operations. Although separate considerations of these two issues simplify the system design, it is often the non-optimal lifetime expectancy for wireless sensor networks. This paper proposes an integral framework that integrates these two correlated items in an interactive entirety. For that, we develop the clustering problems using nonlinear programming. Evolution process of clustering is provided in simulations. Results show that our joint-design proposal reaches the near optimal match between member nodes and cluster heads.     

Computer networks design using hybrid fuzzy expert systems

 Designing and configuring large computer networks to support a variety of applications and computational environments is difficult, as it not only requires highly specialized technical skills and knowledge, but also a deep understanding of a dynamic commercial market. Hybrid fuzzy expert systems integrate fuzzy expert systems and neural networks methods replacing classical hard decision methods and providing better performance than traditional techniques. In this paper, we present an integrated fuzzy expert system, machine learning, and neural networks approach to large structured computer networks design and evaluation. After presenting an overview of the system and the major research choices, we describe in detail the system's modules and present examples of its potential use.     

Solving Electrical Distribution Problems Using Hybrid Evolutionary Data Analysis Techniques

Real-world electrical engineering problems can take advantage of the last Data Analysis methodologies. In this paper we will show that Genetic Fuzzy Rule-Based Systems and Genetic Programming techniques are good choices for tackling with some practical modeling problems. We claim that both evolutionary processes may produce good numerical results while providing us with a model that can be interpreted by a human being. We will analyze in detail the characteristics of these two methods and we will compare them to the some of the most popular classical statistical modeling methods and neural networks.     

A novel application of artificial neural networks to structural analysis

In this work we use the continuous Hopfield network and the continuous bidirectional associative memory system (BAM) in order to develop two novel methods for structural analysis. The development of these techniques is based on the analogous relationship that results from comparing the energy functions of the above two models with that of the structural displacement method (i.e. the socalled stiffness matrix method) and it takes advantage of the fact that classical numerical methods do not have the characteristics of parallel computation that artificial neural networks have. Several examples related to structural deformation are used to illustrate the superiority of the BAM-based neural networks over other traditional numerical methods and the Hopfield model, especially for the case of large dimensional stiffness matrices.     

Training algorithms for Radial Basis Function Networks to tackle learning processes with imbalanced data-sets

Nowadays, many real applications comprise data-sets where the distribution of the classes is significantly different. These data-sets are commonly known as imbalanced data-sets. Traditional classifiers are not able to deal with these kinds of data-sets because they tend to classify only majority classes, obtaining poor results for minority classes. The approaches that have been proposed to address this problem can be categorized into three types: resampling methods, algorithmic adaptations and cost sensitive techniques.Radial Basis Function Networks (RBFNs), artificial neural networks composed of local models or RBFs, have demonstrated their efficiency in different machine learning areas. Centers, widths and output weights for the RBFs must be determined when designing RBFNs.Taking into account the locally tuned response of RBFs, the objective of this paper is to study the influence of global and local paradigms on the weights training phase, within the RBFNs design methodology, for imbalanced data-sets. Least Mean Square and the Singular Value Decomposition have been chosen as representatives of local and global weights training paradigms respectively. These learning algorithms are inserted into classical RBFN design methods that are run on imbalanced data-sets and also on these data-sets preprocessed with re-balance techniques. After applying statistical tests to the results obtained, some guidelines about the RBFN design methodology for imbalanced data-sets are provided.     

Metamodels for Computer-based Engineering Design: Survey and recommendations

The use of statistical techniques to build approximations of expensive computer analysis codes pervades much of today’s engineering design. These statistical approximations, or metamodels, are used to replace the actual expensive computer analyses, facilitating multidisciplinary, multiobjective optimization and concept exploration. In this paper, we review several of these techniques, including design of experiments, response surface methodology, Taguchi methods, neural networks, inductive learning and kriging. We survey their existing application in engineering design, and then address the dangers of applying traditional statistical techniques to approximate deterministic computer analysis codes. We conclude with recommendations for the appropriate use of statistical approximation techniques in given situations, and how common pitfalls can be avoided.     

Minimum-weight spanning tree algorithms A survey and empirical study

The minimum-weight spanning tree problem is one of the most typical and well-known problems of combinatorial optimisation. Efficient solution techniques had been known for many years. However, in the last two decades asymptotically faster algorithms have been invented. Each new algorithm brought the time bound one step closer to linearity and finally Karger, Klein and Tarjan proposed the only known expected linear-time method. Modern algorithms make use of more advanced data structures and appear to be more complicated to implement. Most authors and practitioners refer to these but still use the classical ones, which are considerably simpler but asymptotically slower. The paper first presents a survey of the classical methods and the more recent algorithmic developments. Modern algorithms are then compared with the classical ones and their relative performance is evaluated through extensive empirical tests, using reasonably large-size problem instances. Randomly generated problem instances used in the tests range from small networks having 512 nodes and 1024 edges to quite large ones with 16 384 nodes and 524 288 edges. The purpose of the comparative study is to investigate the conjecture that modern algorithms are also easy to apply and have constants of proportionality small enough to make them competitive in practice with the older ones.Scope and purposeThe minimum-weight spanning tree (MST) problem is a well-known combinatorial optimisation problem concerned with finding a spanning tree of an undirected, connected graph, such that the sum of the weights of the selected edges is minimum. The importance of this problem derives from its direct applications in the design of computer, communication, transportation, power and piping networks; from its appearance as part of solution methods to other problems to which it applies less directly such as network reliability, clustering and classification problems and from its occurrence as a subproblem in the solution of other problems like the travelling salesman problem, the multi-terminal flow problem, the matching problem and the capacitated MST problem. Although efficient solution techniques capable of solving large instances had existed, there has been sustained effort over the last two decades to invent asymptotically faster algorithms. With each new algorithm found the time bound approached linearity. Finally, an expected linear-time method has been proposed. The purpose of this work is to survey the classical and modern solution techniques and empirically compare the performance of the existing methods.     

Bayesian networks for imputation in classification problems

Missing values are an important problem in data mining. In order to tackle this problem in classification tasks, we propose two imputation methods based on Bayesian networks. These methods are evaluated in the context of both prediction and classification tasks. We compare the obtained results with those achieved by classical imputation methods (Expectation–Maximization, Data Augmentation, Decision Trees, and Mean/Mode). Our simulations were performed by means of four datasets (Congressional Voting Records, Mushroom, Wisconsin Breast Cancer and Adult), which are benchmarks for data mining methods. Missing values were simulated in these datasets by means of the elimination of some known values. Thus, it is possible to assess the prediction capability of an imputation method, comparing the original values with the imputed ones. In addition, we propose a methodology to estimate the bias inserted by imputation methods in classification tasks. In this sense, we use four classifiers (One Rule, Naïve Bayes, J4.8 Decision Tree and PART) to evaluate the employed imputation methods in classification scenarios. Computing times consumed to perform imputations are also reported. Simulation results in terms of prediction, classification, and computing times allow us performing several analyses, leading to interesting conclusions. Bayesian networks have shown to be competitive with classical imputation methods.     

无线传感器网络节点定位系统与算法的研究和发展

介绍了无线传感器网络节点定位的概念和原理，节点定位计算的一般过程．讨论了传感器网络各种定位系统常用的测距技术与节点定位计算的基本方法，对几种典型的定位算法进行了定性比较．介绍了节点定位系统与算法的最新发展，对传感器网络节点定位系统和算法进行了必要的总结和概括．     

Elastic analysis of pin joints

The advent of large and fast digital computers and development of numerical techniques suited to these have made it possible to review the analysis of important fundamental and practical problems and phenomena of engineering which have remained intractable for a long time. The understanding of the load transfer between pin and plate is one such. Inspite of continuous attack on these problems for over half a century, classical solutions have remained limited in their approach and value to the understanding of the phenomena and the generation of design data. On the other hand, the finite element methods that have grown simultaneously with the recent development of computers have been helpful in analysing specific problems and answering specific questions, but are yet to be harnessed to assist in obtaining with economy a clearer understanding of the phenomena of partial separation and contact, friction and slip, and fretting and fatigue in pin joints. Against this background, it is useful to explore the application of the classical simple differential equation methods with the aid of computer power to open up this very important area. In this paper we describe some of the recent and current work at the Indian Institute of Science in this last direction.     

Intelligent sensory evaluation: Concepts,implementations, and applications

Sensory evaluation has been widely applied in different industrial fields especially for quality inspection, product design and marketing. Classically, factorial multivariate methods are the only tool for analyzing and modeling sensory data provided by experts, panelists or consumers. These methods are efficient for solving some problems but sometimes cause important information lost. In this situation, new methods based on intelligent techniques such as fuzzy logic, neural networks, data aggregation, classification, clustering have been applied for solving uncertainty and imprecision related to sensory evaluation. These new methods can be used together with the classical ones in a complementary way for obtaining relevant information from sensory data. This paper outlines the general background of sensory evaluation and the corresponding industrial interests and explicitly indicates some orientations for further development by IT researchers.     

Intrusion Detection Systems in Internet of Things and Mobile Ad-Hoc Networks

Internet of Things (IoT) devices work mainly in wireless mediums; requiring different Intrusion Detection System (IDS) kind of solutions to leverage 802.11 header information for intrusion detection. Wireless-specific traffic features with high information gain are primarily found in data link layers rather than application layers in wired networks. This survey investigates some of the complexities and challenges in deploying wireless IDS in terms of data collection methods, IDS techniques, IDS placement strategies, and traffic data analysis techniques. This paper’s main finding highlights the lack of available network traces for training modern machine-learning models against IoT specific intrusions. Specifically, the Knowledge Discovery in Databases (KDD) Cup dataset is reviewed to highlight the design challenges of wireless intrusion detection based on current data attributes and proposed several guidelines to future-proof following traffic capture methods in the wireless network (WN). The paper starts with a review of various intrusion detection techniques, data collection methods and placement methods. The main goal of this paper is to study the design challenges of deploying intrusion detection system in a wireless environment. Intrusion detection system deployment in a wireless environment is not as straightforward as in the wired network environment due to the architectural complexities. So this paper reviews the traditional wired intrusion detection deployment methods and discusses how these techniques could be adopted into the wireless environment and also highlights the design challenges in the wireless environment. The main wireless environments to look into would be Wireless Sensor Networks (WSN), Mobile Ad Hoc Networks (MANET) and IoT as this are the future trends and a lot of attacks have been targeted into these networks. So it is very crucial to design an IDS specifically to target on the wireless networks.     

A memetic algorithm applied to the design of water distribution networks

The optimal design of water distribution networks is a real optimization problem that consists of finding the best way to convey water from the sources to the users, satisfying their requirements. Many researchers have reported algorithms for minimizing the network cost applying a large variety of techniques, such as linear programming, non-linear programming, global optimization methods and meta-heuristic approaches. However, a totally satisfactory and efficient method is not available as yet. Many works have assessed the performance of these techniques using small or medium-sized benchmark networks proposed in the literature, but few of them have tested these methods with large-scale real networks. This paper introduces a new memetic algorithm for the optimal design of water distribution networks. In order to establish an accurate conclusion, five other approaches have also been adapted, namely simulated annealing, mixed simulated annealing and tabu search, scatter search, genetic algorithms and binary linear integer programming. The results obtained in three water distribution networks show that the memetic algorithm performs better than the other methods, especially when the size of the problem increases.     

无线传感器网络位置辅助路由算法研究*

介绍了位置辅助路由的设计特点和准则,分析并定性比较了现有位置辅助路由的贪婪转发策略以及遇到路由空洞时的各类恢复策略,最后指出了需要进一步研究的方向.     

A survey of graph-modification techniques for privacy-preserving on networks

Recently, a huge amount of social networks have been made publicly available. In parallel, several definitions and methods have been proposed to protect users’ privacy when publicly releasing these data. Some of them were picked out from relational dataset anonymization techniques, which are riper than network anonymization techniques. In this paper we summarize privacy-preserving techniques, focusing on graph-modification methods which alter graph’s structure and release the entire anonymous network. These methods allow researchers and third-parties to apply all graph-mining processes on anonymous data, from local to global knowledge extraction.     

神经网络在退化图像复原领域的进展综述

退化图像复原是图像计算领域中的一个重要的难题。近年来以深度学习为代表的 人工智能(AI)技术取得了快速的发展,越来越多的基于神经网络解决退化图像复原的研究工作 出现。首先介绍了神经网络进行退化图像还原的主要技术并对图像复原的问题进行分类;然后 利用神经网络解决退化图像复原问题中细分的多个主要问题,并对每个问题的当前研究现状与 多种基于深度学习网络的解决方法的优势与局限性进行归纳分析,并给出与传统方法的对比。 最后介绍了基于对抗神经网络的极限退化图像复原的新方法,并对未来前景进行展望。     

融合语言特征的卷积神经网络的反讽识别方法

面对某些热点事件,微博评论者经常使用反讽来表达对于该事件的看法,以往的情感分析任务往往忽略这一语言现象。为了提高微博情感分析的准确率,该文对反讽识别开展了研究。通过分析中文文本的语言现象和社交网络的特性,归纳了中文微博反讽的语言特征,提出了一种融合语言特征的卷积神经网络(CNN)的反讽识别方法。该方法将反讽特征和句子分别采用Word Embedding作为输入,再卷积、池化后,将其全连接融合,构建了新的卷积神经网络模型。实验结果表明,该方法在反讽识别的性能上优于传统的基于机器学习的方法。