Improving the Efficiency of Large-Scale Agent-Based Models Using Compression Techniques

Agent-based modelling is becoming a widely used approach for simulating complex phenomena. By making use of emergent behaviour, agent based models can simulate systems right down to the most minute interactions that affect a system’s behaviour. In order to capture the level of detail desired by users, many agent based models now contain hundreds of thousands and even millions of interacting agents. The scale of these models makes them computationally expensive to operate in terms of memory and CPU time, limiting their practicality and use. This chapter details the techniques for applying Dynamic Hierarchical Agent Compression to agent based modelling systems, with the aim of reducing the amount of memory and number of CPU cycles required to manage a set of agents within a model. The scheme outlined extracts the state data stored within a model’s agents and takes advantage of redundancy in this data to reduce the memory required to represent this information. The techniques show how a hierarchical data structure can be used to achieve compression of this data and the techniques for implementing this type of structure within an existing modelling system. The chapter includes a case study that outlines the practical considerations related to the application of this scheme to Australia’s National Model for Emerging Livestock Disease Threats that is currently being developed.

Optimisation of Radiation Beam in Linear Nodes Array of Wireless Sensor Network for Improved Performance

This chapter presents an intelligent method of optimising the radiation beam of wireless sensor nodes in Wireless Sensor Network (WSN). Each node has the feature of a monopole antenna. The optimisation involves selection of nodes to be organised as close as possible to a uniform linear array (ULA) in order to minimise the position errors, which will improve the radiation beam reconfiguring performance. Instead of utilising random beamforming, which needs a large number of sensor nodes to interact with each other and form a narrow radiation beam, the developed optimisation algorithm is emphasized to only a selected number of sensor nodes which can construct a linear array. Thus, the method utilises radiation beam reconfiguration technique to intelligently establish a communication link in a WSN.

On the Use of Particle Swarm Optimization Techniques for Channel Assignments in Cognitive Radio Networks

Efficient utilization of open spectrum in cognitive radio networks requires appropriate allocation of idle spectrum frequency bands (not used by licensed users) among coexisting cognitive radios (secondary users) while minimizing interference among all users. This problem is referred to as the spectrum allocation or the channel assignment problem in cognitive radio networks, and is shown to be NP-hard. Accordingly, different optimization techniques based on evolutionary algorithms were needed in order to solve the channel assignment problem. This chapter investigates the use of particular swarm optimization (PSO) techniques to solve the channel assignment problem in cognitive radio networks. In particular, the authors study the definitiveness of using the native PSO algorithm and the Improved Binary PSO (IBPSO) algorithm to solve the assignment problem. In addition, the performance of these algorithms is compared to that of a fine-tuned genetic algorithm (GA) for this particular problem. Three utilization functions, namely, Mean-Reward, Max-Min-Reward, and Max-Proportional-Fair, are used to evaluate the effectiveness of three optimization algorithms. Extensive simulation results show that PSO and IBPSO algorithms outperform that fine-tuned GA. More interestingly, the native PSO algorithm outperforms both the GA and the IBPSO algorithms in terms of solution speed and quality.

Using Computational Intelligence for Improved Teleoperation over Wireless Network

The role of computational intelligence techniques in applied sciences and engineering is becoming popular today. It is essential because the autonomous engineering applications require intelligent decision in real time in order to achieve the desired goal. This chapter discusses some of the approaches to demonstrate various applications of computational intelligence in dependable networked control systems and a case study of teleoperation over wireless network. The results have shown that computational intelligence algorithms can be successfully implemented on an embedded application to offer an improved online performance. The different approaches have been compared and could be chosen as per application requirements.

Efficient Pronunciation Assessment of Taiwanese-Accented English Based on Unsupervised Model Adaptation and Dynamic Sentence Selection

This chapter presents an efficient approach to personalized pronunciation assessment of Taiwanese-accented English. The main goal of this study is to detect frequently occurring mispronunciation patterns of Taiwanese-accented English instead of scoring English pronunciations directly. The proposed assessment help quickly discover personalized mispronunciations of a student, thus English teachers can spend more time on teaching or rectifying students’ pronunciations. In this approach, an unsupervised model adaptation method is performed on the universal acoustic models to recognize the speech of a specific speaker with mispronunciations and Taiwanese accent. A dynamic sentence selection algorithm, considering the mutual information of the related mispronunciations, is proposed to choose a sentence containing the most undetected mispronunciations in order to quickly extract personalized mispronunciations. The experimental results show that the proposed unsupervised adaptation approach obtains an accuracy improvement of about 2.1% on the recognition of Taiwanese-accented English speech.

Embedded System for Heart Disease Recognition using Fuzzy Clustering and Correlation

This chapter presents the viability analysis and the development of heart disease identification embedded system. It offers a time reduction on electrocardiogram – ECG signal processing by reducing the amount of data samples without any significant loss. The goal of the developed system is the analysis of heart signals. The ECG signals are applied into the system that performs an initial filtering, and then uses a Gustafson-Kessel fuzzy clustering algorithm for the signal classification and correlation. The classification indicates common heart diseases such as angina, myocardial infarction and coronary artery diseases. The system uses the European electrocardiogram ST-T Database – EDB as a reference for tests and evaluation. The results prove the system can perform the heart disease detection on a data set reduced from 213 to just 20 samples, thus providing a reduction to just 9.4% of the original set, while maintaining the same effectiveness. This system is validated in a Xilinx Spartan®-3A FPGA. The FPGA implemented a Xilinx Microblaze® Soft-Core Processor running at a 50 MHz clock rate.

A Multilevel Thresholding Method Based on Multiobjective Optimization for Non-Supervised Image Segmentation

The aim of this work is to provide a comprehensive review of multiobjective optimization in the image segmentation problem based on image thresholding. The authors show that the inclusion of several criteria in the thresholding segmentation process helps to overcome the weaknesses of these criteria when used separately. In this context, they give a recent literature review, and present a new multi-level image thresholding technique, called Automatic Threshold, based on Multiobjective Optimization (ATMO). That combines the flexibility of multiobjective fitness functions with the power of a Binary Particle Swarm Optimization algorithm (BPSO), for searching the “optimum” number of the thresholds and simultaneously the optimal thresholds of three criteria: the between-class variances criterion, the minimum error criterion and the entropy criterion. Some examples of test images are presented to compare with this segmentation method, based on the multiobjective optimization approach with Otsu’s, Kapur’s, and Kittler’s methods. Experimental results show that the thresholding method based on multiobjective optimization is more efficient than the classical Otsu’s, Kapur’s, and Kittler’s methods.

Bangla Music Genre Classification

Bangladesh is very rich in its musical history. Music documented the lives of the people from the ancient times. This chapter provides a guideline for classifying Bangla songs into different genres using a machine learning approach. Four different genres, namely Rabindrasangit, Folk song, Adhunik song, and Pop music, were used in the experiments. A set of second order features are used for representing the trend of change of primary features computed over the timeline of the song. The features are incorporated into a number of classification algorithms and a classification framework is developed. The uniqueness of the genres is clearly revealed by high classification accuracies achieved by the different classifiers.

Which Fundamental Factors Proxy for Share Returns?

It is widely accepted that the presence of some of the firm’s attributes or characteristics attracting premiums in terms of average returns is pervasive and not restricted to a few individual markets. However, the way to derive these premiums by sorting firms based on their characteristics that are known associated with share returns is not without controversy. This chapter takes a different approach by adopting a novel Multi Self-Organising Maps to cluster shares first and then identify fundamental factors afterwards. It finds that firm’s size and book-to-market ratio attributes do have explanatory power over share average returns. There is also lack of evidence for other factors in explaining the share average returns.

Fuzzy Image Segmentation for Mass Detection in Digital Mammography

In the last decade, many computer-aided diagnosis (CAD) systems that utilize a broad range of diagnostic techniques have been proposed. Due to both the inherently complex structure of the breast tissues and the low intensity contrast found in most mammographic images, CAD systems that are based on conventional techniques have been shown to have missed malignant masses in mammographic images that would otherwise be treatable. On the other hand, systems based on fuzzy image processing techniques have been found to be able to detect masses in cases where conventional techniques would have failed. In the current chapter, recent advances in fuzzy image segmentation techniques as applied to mass detection in digital mammography are reviewed. Image segmentation is an important step in CAD systems since the quality of its outcome will significantly affect the processing downstream that can involve both detection and classification of benign versus malignant masses.