Computationally efficient sequential learning algorithms for direct link resource-allocating networks

Computionally efficient sequential learning algorithms are developed for direct-link resource-allocating networks (DRANs). These are achieved by decomposing existing recursive training algorithms on a layer by layer and neuron by neuron basis. This allows network weights to be updated in an efficient parallel manner and facilitates the implementation of minimal update extensions that yield a significant reduction in computation load per iteration compared to existing sequential learning methods employed in resource-allocation network (RAN) and minimal RAN (MRAN) approaches. The new algorithms, which also incorporate a pruning strategy to control network growth, are evaluated on three different system identification benchmark problems and shown to outperform existing methods both in terms of training error convergence and computational efficiency. (c) 2005 Elsevier B.V. All rights reserved.

A Capabilities-Aware Framework for Using Computational Accelerators in Data-Intensive Computing

Multicore computational accelerators such as GPUs are now commodity components for highperformance computing at scale. While such accelerators have been studied in some detail as stand-alone computational engines, their integration in large-scale distributed systems raises new challenges and trade-offs. In this paper, we present an exploration of resource management alternatives for building asymmetric accelerator-based distributed systems. We present these alternatives in the context of a capabilities-aware framework for data-intensive computing, which uses an enhanced implementation of the MapReduce programming model for accelerator-based clusters, compared to the state of the art. The framework can transparently utilize heterogeneous accelerators for deriving high performance with low programming effort. Our work is the first to compare heterogeneous types of accelerators, GPUs and a Cell processors, in the same environment and the first to explore the trade-offs between compute-efficient and control-efficient accelerators on data-intensive systems. Our investigation shows that our framework scales well with the number of different compute nodes. Furthermore, it runs simultaneously on two different types of accelerators, successfully adapts to the resource capabilities, and performs 26.9% better on average than a static execution approach.

Exploiting Human Resource Requirements to Infer Human Movement Patterns for Use in Modelling Disease Transmission Systems: An Example from Eastern Province, Zambia

In this research, an agent-based model (ABM) was developed to generate human movement routes between homes and water resources in a rural setting, given commonly available geospatial datasets on population distribution, land cover and landscape resources. ABMs are an object-oriented computational approach to modelling a system, focusing on the interactions of autonomous agents, and aiming to assess the impact of these agents and their interactions on the system as a whole. An A* pathfinding algorithm was implemented to produce walking routes, given data on the terrain in the area. A* is an extension of Dijkstra's algorithm with an enhanced time performance through the use of heuristics. In this example, it was possible to impute daily activity movement patterns to the water resource for all villages in a 75 km long study transect across the Luangwa Valley, Zambia, and the simulated human movements were statistically similar to empirical observations on travel times to the water resource (Chi-squared, 95% confidence interval). This indicates that it is possible to produce realistic data regarding human movements without costly measurement as is commonly achieved, for example, through GPS, or retrospective or real-time diaries. The approach is transferable between different geographical locations, and the product can be useful in providing an insight into human movement patterns, and therefore has use in many human exposure-related applications, specifically epidemiological research in rural areas, where spatial heterogeneity in the disease landscape, and space-time proximity of individuals, can play a crucial role in disease spread.