Data-Intensive Computing in the 21st Century

The deluge of data that future applications must process—in domains ranging from science to business informatics—creates a compelling argument for substantially increased R&D targeted at discovering scalable hardware and software solutions for data-intensive problems.     

Eigenfeature regularization and extraction in face recognition

This work proposes a subspace approach that regularizes and extracts eigenfeatures from the face image. Eigenspace of the within-class scatter matrix is decomposed into three subspaces: a reliable subspace spanned mainly by the facial variation, an unstable subspace due to noise and finite number of training samples and a null subspace. Eigenfeatures are regularized differently in these three subspaces based on an eigenspectrum model to alleviate problems of instability, over-fitting or poor generalization. This also enables the discriminant evaluation performed in the whole space. Feature extraction or dimensionality reduction occurs only at the final stage after the discriminant assessment. These efforts facilitate a discriminative and stable low-dimensional feature representation of the face image. Experiments comparing the proposed approach with some other popular subspace methods on the FERET, ORL, AR and GT databases show that our method consistently outperforms others.     

A two-stage hardware scheduler combining greedy and optimal scheduling

Greedy scheduling heuristics provide a low complexity and scalable albeit particularly sub-optimal strategy for hardware-based crossbar schedulers. In contrast, the maximum matching algorithm for Bipartite graphs can be used to provide optimal scheduling for crossbar-based interconnection networks with a significant complexity and scalability cost. In this paper, we show how maximum matching can be reformulated in terms of Boolean operations rather than the more traditional formulations. By leveraging the inherent parallelism available in custom hardware design, we reformulate maximum matching in terms of Boolean operations rather than matrix computations and introduce three maximum matching implementations in hardware. Specifically, we examine a Pure Logic Scheduler with three dimensions of parallelism, a Matrix Scheduler with two dimensions of parallelism and a Vector Scheduler with one dimension of parallelism. These designs reduce the algorithmic complexity for an N×N$N\times N$ network from O(N³)$O\left(N^3\right)$ to O(1)$O\left(1\right)$, O(K)$O\left(K\right)$, and O(KN)$O\left(KN\right)$, respectively, where K$K$ is the number of optimization steps. While an optimal scheduling algorithm requires K=2N−1$K=2N-1$ steps, by starting with our hardware-based greedy strategy to generate an initial schedule, our simulation results show that the maximum matching scheduler can achieve 99% of the optimal schedule when K=9$K=9$. We examine hardware and time complexity of these architectures for crossbar sizes of up to N=1024$N=1024$. Using FPGA synthesis results, we show that a greedy schedule for crossbars, ranging from 8×8 to 256×256, can be optimized in less than 20 ns per optimization step. For crossbars reaching 1024×1024 the scheduling can be completed in approximately 10 μs with current technology and could reach under 90 ns with future technologies.     

Incremental application of knowledge to continuously arriving time-oriented data

In our previous work, we introduced a computational architecture that effectively supports the tasks of continuous monitoring and of aggregation querying of complex domain meaningful time-oriented concepts and patterns (temporal abstractions), in environments featuring large volumes of continuously arriving and accumulating time-oriented raw data. Examples include provision of decision support in clinical medicine, making financial decisions, detecting anomalies and potential threats in communication networks, integrating intelligence information from multiple sources, etc. In this paper, we describe the general, domain-independent but task-specific problem-solving method underling our computational architecture, which we refer to as incremental knowledge-based temporal abstraction (IKBTA). The IKBTA method incrementally computes temporal abstractions by maintaining persistence and validity of continuously computed temporal abstractions from arriving time-stamped data. We focus on the computational framework underlying our reasoning method, provide well-defined semantic and knowledge requirements for incremental inference, which utilizes a logical model of time, data, and high-level abstract concepts, and provide a detailed analysis of the computational complexity of our approach.     

Scan Cell Positioning for Boosting the Compression of Fan-Out Networks

Ensuring a high manufacturing test quality of an integrated electronic circuit mandates the application of a large volume test set. Even if the test data can be fit into the memory of an external tester, the consequent increase in test application time reflects into elevated production costs. Test data compression solutions have been proposed to address the test time and data volume problem by storing and delivering the test data in a compressed format, and subsequently by expanding the data on-chip. In this paper, we propose a scan cell positioning methodology that accompanies a compression technique in order to boost the compression ratio, and squash the test data even further. While we present the application of the proposed approach in conjunction with the fan-out based decompression architecture, this approach can be extended for application along with other compression solutions as well. The experimental results also confirm the compression enhancement of the proposed methodology.     

Perceived interactivity leading to e-loyalty: Development of a model for cognitive–affective user responses

Novel applications of website interactivity are important to attract and retain online users. In this empirical study five designs for interactivity are examined using different web-poll interfaces. The goal of the investigation is to examine perceived interactivity in a model which includes most commonly tested cognitive elements such as efficiency and effectiveness, but augments this model with the inclusion of a cognitive–affective element for trust, and an affective element of enjoyment. More specifically, a model is created to validate the relationship of perceived interactivity (comprised of user control, user connectedness, and responsiveness of the web-poll application) to efficiency, effectiveness, trust and enjoyment, of the website. In turn, efficiency, effectiveness, trust, and enjoyment are tested for their influence on user behavioral intentions for e-loyalty. All relationships in the model are supported. In addition, exploratory evaluation of qualitative comments is conducted to investigate additional insights between the five web-poll treatments in this investigation. The research confirms the complexity of a model in which cognitive, cognitive–affective and affective elements are present, and advances knowledge on the consequences of perceived interactivity. In additional to theoretical advancements, the research has merit for web designers and online marketers regarding how to enhance interactive online web applications.     

Delay- and Disruption-Tolerant Networking

Delay- and disruption-tolerant networking (DTN) grew out of attempts to develop an Interplanetary Internet but has evolved into an active area of networking research, with applications in space networking, military tactical networking, and networking for various challenged communities. The DTN Research Group provides an open forum in which DTN researchers and developers can collaborate to further develop this experimental technology.     

Hierarchical classification of protein function with ensembles of rules and particle swarm optimisation

This paper focuses on hierarchical classification problems where the classes to be predicted are organized in the form of a tree. The standard top-down divide and conquer approach for hierarchical classification consists of building a hierarchy of classifiers where a classifier is built for each internal (non-leaf) node in the class tree. Each classifier discriminates only between its child classes. After the tree of classifiers is built, the system uses them to classify test examples one class level at a time, so that when the example is assigned a class at a given level, only the child classes need to be considered at the next level. This approach has the drawback that, if a test example is misclassified at a certain class level, it will be misclassified at deeper levels too. In this paper we propose hierarchical classification methods to mitigate this drawback. More precisely, we propose a method called hierarchical ensemble of hierarchical rule sets (HEHRS), where different ensembles are built at different levels in the class tree and each ensemble consists of different rule sets built from training examples at different levels of the class tree. We also use a particle swarm optimisation (PSO) algorithm to optimise the rule weights used by HEHRS to combine the predictions of different rules into a class to be assigned to a given test example. In addition, we propose a variant of a method to mitigate the aforementioned drawback of top-down classification. These three types of methods are compared against the standard top-down hierarchical classification method in six challenging bioinformatics datasets, involving the prediction of protein function. Overall HEHRS with the rule weights optimised by the PSO algorithm obtains the best predictive accuracy out of the four types of hierarchical classification method.     

Agent Technologies for Sensor Networks

Wireless sensor networks are increasingly seen as a solution to the problem of performing continuous wide-area monitoring in many environmental, security, and military scenarios. The distributed nature of such networks and the autonomous behavior expected of them present many novel challenges. In this article, the authors argue that a new synthesis of electronic engineering and agent technology is required to address these challenges, and they describe three examples where this synthesis has succeeded. In more detail, they describe how these novel approaches address the need for communication and computationally efficient decentralized algorithms to coordinate the behavior of physically distributed sensors, how they enable the real-world deployment of sensor agent platforms in the field, and finally, how they facilitate the development of intelligent agents that can autonomously acquire data from these networks and perform information processing tasks such as fusion, inference, and prediction.     

USE of Railway Analysis Tools from an Australian Perspective

The functions of Australia’s railways are divided between the delivery of suburban passenger railway services, long distance general and intermodal freight services and regional bulk commodity haulage. There is a national interest in the efficient carriage of export freight flows because Australia is a major trading nation. Australia’s railways are expected play their part in hauling export commodities. Unsurprisingly, there are conflicting demands on these railways. Hence, universities, and the industry itself, are directing research and investigation effort into policy, planning, engineering, operational and human factors matters. A question which then arises is what means could be used to analyse the problems identified during this research and investigation. This paper is thus concerned with the inter-relationships between the planning, engineering and operations of railways in Australia. It identifies four areas of analysis associated with the planning and development of railway infrastructure and operations. It then discusses a range of analytical tools which could be applied to different components of these analytical areas and critiques their appropriateness from an Australian perspective. Having made this assessment, the paper uses a recent Australian case study to show how analytical tools could be used and what lessons might be learnt from the process. Figure Use of Railway Analysis Tools from an Australian Perspective