A constructive learning algorithm based on back-propagation

There are been a resurgence of interest in the neural networks field in recent years, provoked in part by the discovery of the properties of multi-layer networks. This interest has in turn raised questions about the possibility of making neural network behaviour more adaptive by automating some of the processes involved. Prior to these particular questions, the process of determining the parameters and network architecture required to solve a given problem had been a time consuming activity. A number of researchers have attempted to address these issues by automating these processes, concentrating in particular on the dynamic selection of an appropriate network architecture.The work presented here specifically explores the area of automatic architecture selection; it focuses upon the design and implementation of a dynamic algorithm based on the Back-Propagation learning algorithm. The algorithm constructs a single hidden layer as the learning process proceeds using individual pattern error as the basis of unit insertion. This algorithm is applied to several problems of differing type and complexity and is found to produce near minimal architectures that are shown to have a high level of generalisation ability.

Multi-function controllers for a hydraulic pump

This project has been undertaken for Hamworthy Hydraulics Limited. Its objective was to design and develop a controller package for a variable displacement, hydraulic pump for use mainly on mobile earth moving machinery. A survey was undertaken of control options used in practice and from this a design specification was formulated, the successful implementation of which would give Hamworthy an advantage over its competitors. Two different modes for the controller were envisaged. One consisted of using conventional hydro-mechanics and the other was based upon a microprocessor. To meet short term customer prototype requirements the first section of work was the realisation of the hydro-mechanical system. Mathematical models were made to evaluate controller stability and hence aid their design. The final package met the requirements of the specification and a single version could operate all sizes of variable displacement pumps in the Hamworthy range. The choice of controller options and combinations totalled twenty-four. The hydro-mechanical controller was complex and it was realised that a micro-processor system would allow all options to be implemented with just one design of hardware, thus greatly simplifying production. The final section of this project was to determine whether such a design was feasible. This entailed finding cheap, reliable transducers, using mathematical models to predict electro-hydraulic interface stability, testing such interfaces and finally incorporating a micro-processor in an interactive control loop. The study revealed that such a system was technically possible but it would cost 60% more than its hydro-mechanical counterpart. It was therefore concluded that, in the short term, for the markets considered, the hydro-mechanical design was the better solution. Regarding the micro-processor system the final conclusion was that, because the relative costs of the two systems are decreasing, the electro-hydraulic controller will gradually become more attractive and therefore Hamworthy should continue with its development.

The integration of multi-source data to improve the classification of remotely sensed images

The number of remote sensing platforms and sensors rises almost every year, yet much work on the interpretation of land cover is still carried out using either single images or images from the same source taken at different dates. Two questions could be asked of this proliferation of images: can the information contained in different scenes be used to improve the classification accuracy and, what is the best way to combine the different imagery? Two of these multiple image sources are MODIS on the Terra platform and ETM+ on board Landsat7, which are suitably complementary. Daily MODIS images with 36 spectral bands in 250-1000 m spatial resolution and seven spectral bands of ETM+ with 30m and 16 days spatial and temporal resolution respectively are available. In the UK, cloud cover may mean that only a few ETM+ scenes may be available for any particular year and these may not be at the time of year of most interest. The MODIS data may provide information on land cover over the growing season, such as harvest dates, that is not present in the ETM+ data. Therefore, the primary objective of this work is to develop a methodology for the integration of medium spatial resolution Landsat ETM+ image, with multi-temporal, multi-spectral, low-resolution MODIS \Terra images, with the aim of improving the classification of agricultural land. Additionally other data may also be incorporated such as field boundaries from existing maps. When classifying agricultural land cover of the type seen in the UK, where crops are largely sown in homogenous fields with clear and often mapped boundaries, the classification is greatly improved using the mapped polygons and utilising the classification of the polygon as a whole as an apriori probability in classifying each individual pixel using a Bayesian approach. When dealing with multiple images from different platforms and dates it is highly unlikely that the pixels will be exactly co-registered and these pixels will contain a mixture of different real world land covers. Similarly the different atmospheric conditions prevailing during the different days will mean that the same emission from the ground will give rise to different sensor reception. Therefore, a method is presented with a model of the instantaneous field of view and atmospheric effects to enable different remote sensed data sources to be integrated.

Agent Based Models of Competition and Collaboration

Swarm intelligence is a popular paradigm for algorithm design. Frequently drawing inspiration from natural systems, it assigns simple rules to a set of agents with the aim that, through local interactions, they collectively solve some global problem. Current variants of a popular swarm based optimization algorithm, particle swarm optimization (PSO), are investigated with a focus on premature convergence. A novel variant, dispersive PSO, is proposed to address this problem and is shown to lead to increased robustness and performance compared to current PSO algorithms. A nature inspired decentralised multi-agent algorithm is proposed to solve a constrained problem of distributed task allocation. Agents must collect and process the mail batches, without global knowledge of their environment or communication between agents. New rules for specialisation are proposed and are shown to exhibit improved eciency and exibility compared to existing ones. These new rules are compared with a market based approach to agent control. The eciency (average number of tasks performed), the exibility (ability to react to changes in the environment), and the sensitivity to load (ability to cope with differing demands) are investigated in both static and dynamic environments. A hybrid algorithm combining both approaches, is shown to exhibit improved eciency and robustness. Evolutionary algorithms are employed, both to optimize parameters and to allow the various rules to evolve and compete. We also observe extinction and speciation. In order to interpret algorithm performance we analyse the causes of eciency loss, derive theoretical upper bounds for the eciency, as well as a complete theoretical description of a non-trivial case, and compare these with the experimental results. Motivated by this work we introduce agent "memory" (the possibility for agents to develop preferences for certain cities) and show that not only does it lead to emergent cooperation between agents, but also to a signicant increase in efficiency.

Theoretic analysis of IEEE 802.11 rate adaptation algorithm SampleRate

In this paper we investigate rate adaptation algorithm SampleRate, which spends a fixed time on bit-rates other than the currently measured best bit-rate. A simple but effective analytic model is proposed to study the steady-state behavior of the algorithm. Impacts of link condition, channel congestion and multi-rate retry on the algorithm performance are modeled. Simulations validate the model. It is also observed there is still a large performance gap between SampleRate and optimal scheme in case of high frame collision probability.

Convergence and knowledge processing in multi-agent systems

Multi-agent systems are complex systems comprised of multiple intelligent agents that act either independently or in cooperation with one another. Agent-based modelling is a method for studying complex systems like economies, societies, ecologies etc. Due to their complexity, very often mathematical analysis is limited in its ability to analyse such systems. In this case, agent-based modelling offers a practical, constructive method of analysis. The objective of this book is to shed light on some emergent properties of multi-agent systems. The authors focus their investigation on the effect of knowledge exchange on the convergence of complex, multi-agent systems.

The effects of a lutein-based supplement on objective and subjective measures of retinal and visual function in eyes with age-related maculopathy - a randomised controlled trial

Lutein and zeaxanthin are lipid-soluble antioxidants found within the macula region of the retina. Links have been suggested between increased levels of these carotenoids and reduced risk for age-related macular disease (ARMD). Therefore, the effect of lutein-based supplementation on retinal and visual function in people with early stages of ARMD (age-related maculopathy, ARM) was assessed using multi-focal electroretinography (mfERG), contrast sensitivity and distance visual acuity. A total of fourteen participants were randomly allocated to either receive a lutein-based oral supplement (treated group) or no supplement (non-treated group). There were eight participants aged between 56 and 81 years (65·50 (sd 9·27) years) in the treated group and six participants aged between 61 and 83 years (69·67 (sd 7·52) years) in the non-treated group. Sample sizes provided 80 % power at the 5 % significance level. Participants attended for three visits (0, 20 and 40 weeks). At 60 weeks, the treated group attended a fourth visit following 20 weeks of supplement withdrawal. No changes were seen between the treated and non-treated groups during supplementation. Although not clinically significant, mfERG ring 3 N2 latency (P= 0·041) and ring 4 P1 latency (P= 0·016) increased, and a trend for reduction of mfERG amplitudes was observed in rings 1, 3 and 4 on supplement withdrawal. The statistically significant increase in mfERG latencies and the trend for reduced mfERG amplitudes on withdrawal are encouraging and may suggest a potentially beneficial effect of lutein-based supplementation in ARM-affected eyes. Copyright © 2012 The Authors.

Nonlinear speech analysis algorithms mapped to a standard metric achieve clinically useful quantification of average Parkinson's disease symptom severity

The standard reference clinical score quantifying average Parkinson's disease (PD) symptom severity is the Unified Parkinson's Disease Rating Scale (UPDRS). At present, UPDRS is determined by the subjective clinical evaluation of the patient's ability to adequately cope with a range of tasks. In this study, we extend recent findings that UPDRS can be objectively assessed to clinically useful accuracy using simple, self-administered speech tests, without requiring the patient's physical presence in the clinic. We apply a wide range of known speech signal processing algorithms to a large database (approx. 6000 recordings from 42 PD patients, recruited to a six-month, multi-centre trial) and propose a number of novel, nonlinear signal processing algorithms which reveal pathological characteristics in PD more accurately than existing approaches. Robust feature selection algorithms select the optimal subset of these algorithms, which is fed into non-parametric regression and classification algorithms, mapping the signal processing algorithm outputs to UPDRS. We demonstrate rapid, accurate replication of the UPDRS assessment with clinically useful accuracy (about 2 UPDRS points difference from the clinicians' estimates, p < 0.001). This study supports the viability of frequent, remote, cost-effective, objective, accurate UPDRS telemonitoring based on self-administered speech tests. This technology could facilitate large-scale clinical trials into novel PD treatments.

Clustering web documents using hierarchical representation with multi-granularity

Web document cluster analysis plays an important role in information retrieval by organizing large amounts of documents into a small number of meaningful clusters. Traditional web document clustering is based on the Vector Space Model (VSM), which takes into account only two-level (document and term) knowledge granularity but ignores the bridging paragraph granularity. However, this two-level granularity may lead to unsatisfactory clustering results with “false correlation”. In order to deal with the problem, a Hierarchical Representation Model with Multi-granularity (HRMM), which consists of five-layer representation of data and a twophase clustering process is proposed based on granular computing and article structure theory. To deal with the zero-valued similarity problemresulted from the sparse term-paragraphmatrix, an ontology based strategy and a tolerance-rough-set based strategy are introduced into HRMM. By using granular computing, structural knowledge hidden in documents can be more efficiently and effectively captured in HRMM and thus web document clusters with higher quality can be generated. Extensive experiments show that HRMM, HRMM with tolerancerough-set strategy, and HRMM with ontology all outperform VSM and a representative non VSM-based algorithm, WFP, significantly in terms of the F-Score.

Video-based, real-time multi-view stereo

We investigate the problem of obtaining a dense reconstruction in real-time, from a live video stream. In recent years, multi-view stereo (MVS) has received considerable attention and a number of methods have been proposed. However, most methods operate under the assumption of a relatively sparse set of still images as input and unlimited computation time. Video based MVS has received less attention despite the fact that video sequences offer significant benefits in terms of usability of MVS systems. In this paper we propose a novel video based MVS algorithm that is suitable for real-time, interactive 3d modeling with a hand-held camera. The key idea is a per-pixel, probabilistic depth estimation scheme that updates posterior depth distributions with every new frame. The current implementation is capable of updating 15 million distributions/s. We evaluate the proposed method against the state-of-the-art real-time MVS method and show improvement in terms of accuracy. © 2011 Elsevier B.V. All rights reserved.

Evaluation of hemifield sector analysis protocol in multifocal visual evoked potential (MFVEP) objective perimetry for the diagnosis and early detection of glaucomatous field defects

Visual field assessment is a core component of glaucoma diagnosis and monitoring, and the Standard Automated Perimetry (SAP) test is considered up until this moment, the gold standard of visual field assessment. Although SAP is a subjective assessment and has many pitfalls, it is being constantly used in the diagnosis of visual field loss in glaucoma. Multifocal visual evoked potential (mfVEP) is a newly introduced method used for visual field assessment objectively. Several analysis protocols have been tested to identify early visual field losses in glaucoma patients using the mfVEP technique, some were successful in detection of field defects, which were comparable to the standard SAP visual field assessment, and others were not very informative and needed more adjustment and research work. In this study, we implemented a novel analysis approach and evaluated its validity and whether it could be used effectively for early detection of visual field defects in glaucoma. OBJECTIVES: The purpose of this study is to examine the effectiveness of a new analysis method in the Multi-Focal Visual Evoked Potential (mfVEP) when it is used for the objective assessment of the visual field in glaucoma patients, compared to the gold standard technique. METHODS: 3 groups were tested in this study; normal controls (38 eyes), glaucoma patients (36 eyes) and glaucoma suspect patients (38 eyes). All subjects had a two standard Humphrey visual field HFA test 24-2 and a single mfVEP test undertaken in one session. Analysis of the mfVEP results was done using the new analysis protocol; the Hemifield Sector Analysis HSA protocol. Analysis of the HFA was done using the standard grading system. RESULTS: Analysis of mfVEP results showed that there was a statistically significant difference between the 3 groups in the mean signal to noise ratio SNR (ANOVA p<0.001 with a 95% CI). The difference between superior and inferior hemispheres in all subjects were all statistically significant in the glaucoma patient group 11/11 sectors (t-test p<0.001), partially significant 5/11 (t-test p<0.01) and no statistical difference between most sectors in normal group (only 1/11 was significant) (t-test p<0.9). sensitivity and specificity of the HAS protocol in detecting glaucoma was 97% and 86% respectively, while for glaucoma suspect were 89% and 79%. DISCUSSION: The results showed that the new analysis protocol was able to confirm already existing field defects detected by standard HFA, was able to differentiate between the 3 study groups with a clear distinction between normal and patients with suspected glaucoma; however the distinction between normal and glaucoma patients was especially clear and significant. CONCLUSION: The new HSA protocol used in the mfVEP testing can be used to detect glaucomatous visual field defects in both glaucoma and glaucoma suspect patient. Using this protocol can provide information about focal visual field differences across the horizontal midline, which can be utilized to differentiate between glaucoma and normal subjects. Sensitivity and specificity of the mfVEP test showed very promising results and correlated with other anatomical changes in glaucoma field loss.

Servitization capabilities for advanced services:a multi-actor perspective

Servitization is the process by which manufacturers add services to their product offerings and even replace products with services. The capabilities necessary to develop and deliver advanced services as part of servitization are often discussed in the literature from the manufacturer’s perspective, e.g., having a service-focused culture or the ability to sell solutions. Recent research has acknowledged the important role of customers and, to a lesser extent, other actors (e.g., intermediaries) in bringing about successful servitization, particularly for use-oriented and results-oriented advanced services. The objective of this study is to identify the capabilities required to successful develop advanced services as part of servitization by considering the perspective of manufacturers, intermediaries and customers. This study involved interviews with 33 managers in 28 large UK-based companies from these three groups, about servitization capabilities. The findings suggest that there are eight broad capabilities that are important for advanced services; 1) personnel with expertise and deep technical product knowledge, 2) methodologies for improving operational processes, helping to manage risk and reduce costs, 3) the evolution from being a product- focused manufacturer to embracing a services culture, 4) developing trusting relationships with other actors in the network to support the delivery of advanced services, 5) new innovation activities focused on financing contracts (e.g., ‘gain share’) and technology implementation (e.g., Web-based applications), 6) customer intimacy through understanding their business challenges in order to develop suitable solutions, 7) extensive infrastructure (e.g., personnel, service centres) to deliver a local service, and 8) the ability to tailor service offerings to each customer’s requirements and deliver these responsively to changing needs. The capabilities required to develop and deliver advanced services align to a need to enhance the operational performance of supplied products throughout their lifecycles and as such require greater investment than the capabilities for base and intermediate services.

Using the max-sum algorithm for supply chain emergence in dynamic multiunit environments

Supply chain formation (SCF) is the process of determining the set of participants and exchange relationships within a network with the goal of setting up a supply chain that meets some predefined social objective. Many proposed solutions for the SCF problem rely on centralized computation, which presents a single point of failure and can also lead to problems with scalability. Decentralized techniques that aid supply chain emergence offer a more robust and scalable approach by allowing participants to deliberate between themselves about the structure of the optimal supply chain. Current decentralized supply chain emergence mechanisms are only able to deal with simplistic scenarios in which goods are produced and traded in single units only and without taking into account production capacities or input-output ratios other than 1:1. In this paper, we demonstrate the performance of a graphical inference technique, max-sum loopy belief propagation (LBP), in a complex multiunit unit supply chain emergence scenario which models additional constraints such as production capacities and input-to-output ratios. We also provide results demonstrating the performance of LBP in dynamic environments, where the properties and composition of participants are altered as the algorithm is running. Our results suggest that max-sum LBP produces consistently strong solutions on a variety of network structures in a multiunit problem scenario, and that performance tends not to be affected by on-the-fly changes to the properties or composition of participants.

A diversity dilemma in evolutionary markets

Markets are useful mechanisms for performing resource al- location in fully decentralised computational and other systems, since they can possess a range of desirable properties, such as efficiency, decentralisation, robustness and scalability. In this paper we investigate the behaviour of co-evolving evolutionary market agents as adaptive offer generators for sellers in a multi-attribute posted-offer market. We demonstrate that the evolutionary approach enables sellers to automatically position themselves in market niches, created by heterogeneous buyers. We find that a trade-off exists for the evolutionary sellers between maintaining high population diversity to facilitate movement between niches and low diversity to exploit the current niche and maximise cumulative payoff. We characterise the trade-off from the perspective of the system as a whole, and subsequently from that of an individual seller. Our results highlight a decision on risk aversion for resource providers, but crucially we show that rational self-interested sellers would not adopt the behaviour likely to lead to the ideal result from the system point of view.

Evolutionary Technique of Shorter Route Determination of Fire Brigade Following to Fire Place with the Optimized Space of Search

In this paper the technique of shorter route determination of fire engine to the fire place on time minimization criterion with the use of evolutionary modeling is offered. The algorithm of its realization on the base of complete and optimized space of search of possible decisions is explored. The aspects of goal function forming and program realization of method having a special purpose are considered. Experimental verification is executed and the results of comparative analysis with the expert conclusions are considered.