Dr David Cornforth

© 2017 Elsevier Ltd Social media data can be valuable in many ways. However, the vast amount of content shared and the linguistic variants of languages used on social media are making it very challenging for high-value topics to be identified. In this paper, we present an unsupervised multilingual approach for identifying highly relevant terms and topics from the mass of social media data. This approach combines term ranking, localised language analysis, unsupervised topic clustering and multilingual sentiment analysis to extract prominent topics through analysis of Twitter's tweets from a period of time. It is observed that each of the ranking methods tested has their strengths and weaknesses, and that our proposed ¿Joint¿ ranking method is able to take advantage of the strengths of the ranking methods. This ¿Joint¿ ranking method coupled with an unsupervised topic clustering model is shown to have the potential to discover topics of interest or concern to a local community. Practically, being able to do so may help decision makers to gauge the true opinions or concerns on the ground. Theoretically, the research is significant as it shows how an unsupervised online topic identification approach can be designed without much manual annotation effort, which may have great implications for future development of expert and intelligent systems.

© 2016 Springer Science+Business Media Dordrecht The ability to analyse online user-generated content related to sentiments (e.g., thoughts and opinions) on products or policies has become a de-facto skillset for many companies and organisations. Besides the challenge of understanding formal textual content, it is also necessary to take into consideration the informal and mixed linguistic nature of online social media languages, which are often coupled with localised slang as a way to express ¿true¿ feelings. Due to the multilingual nature of social media data, analysis based on a single official language may carry the risk of not capturing the overall sentiment of online content. While efforts have been made to understand multilingual sentiment analysis based on a range of informal languages, no significant electronic resource has been built for these localised languages. This paper reviews the various current approaches and tools used for multilingual sentiment analysis, identifies challenges along this line of research, and provides several recommendations including a framework that is particularly applicable for dealing with scarce resource languages.

© 2016 Elsevier B.V. All rights reserved. Due to the huge volume and linguistic variation of data shared online, accurate detection of the sentiment of a message (polarity detection) can no longer rely on human assessors or through simple lexicon keyword matching. This paper presents a semi-supervised approach in constructing essential toolkits for analysing the polarity of a localised scarce-resource language, Singlish (Singaporean English). Corpus-based bootstrapping using a multilingual, multifaceted lexicon was applied to construct an annotated testing dataset, while unsupervised methods such as lexicon polarity detection, frequent item extraction through association rules and latent semantic analysis were used to identify the polarity of Singlish n-grams before human assessment was done to isolate misleading terms and remove concept ambiguity. The findings suggest that this multilingual approach outshines polarity analysis using only the English language. In addition, a hybrid combination of the Support Vector Machine and a proposed Singlish Polarity Detection algorithm, which incorporates unigram and n-gram Singlish sentic patterns with other multilingual polarity sentic patterns such as negation and adversative, is able to outperform other approaches in comparison. The promising results of a pooled testing dataset generated from the vast amount of unannotated Singlish data clearly show that our multilingual Singlish sentic pattern approach has the potential to be adopted in real-world polarity detection.

Copyright © 2015 Inderscience Enterprises Ltd. In-match player performance, measured by data from Geographical Positioning System (GPS) devices, was predicted with a correlation coefficient of greater than 0.7. Predictions were based on heart rate variability measures and used advanced regression techniques based on machine learning. These techniques included methods for the selection of variables to be included in the regression study. Results indicate that variable selection using a wrapper subset method with a genetic algorithm outperformed both principal components analysis and the default method of using all variables. The success of prediction of match performance suggests a potential for new tools to assist the team coach in player selection and management of player training. This work also provides the possibility for a training programme to be adjusted specifically to meet the challenges of the size of the playing field and the temperature likely to be encountered on the day of the match.

© 2015 Khalaf, Jelinek, Robinson, Cornforth, Tarvainen and Al-Aubaidy. Background: Physiological interactions are abundant within, and between, body systems. These interactions may evolve into discrete states during pathophysiological processes resulting from common mechanisms. An association between arterial stenosis, identified by low ankle-brachial pressure index (ABPI) and cardiovascular disease (CVD) as been reported. Whether an association between vascular calcification-characterized by high ABPI and a different pathophysiology-is similarly associated with CVD, has not been established. The current study aims to investigate the association between ABPI, and cardiac rhythm, as an indicator of cardiovascular health and functionality, utilizing heart rate variability (HRV). Methods and Results: Two hundred and thirty six patients underwent ABPI assessment. Standard time and frequency domain, and non-linear HRV measures were determined from 5-min electrocardiogram. ABPI data were divided into normal (n = 101), low (n = 67) and high (n = 66) and compared to HRV measures.(DFAa1 and SampEn were significantly different between the low ABPI, high ABPI and control groups (p < 0.05). Conclusion: A possible coupling between arterial stenosis and vascular calcification with decreased and increased HRV respectively was observed. Our results suggest a model for interpreting the relationship between vascular pathophysiology and cardiac rhythm. The cardiovascular system may be viewed as a complex system comprising a number of interacting subsystems. These cardiac and vascular subsystems/networks may be coupled and undergo transitions in response to internal or external perturbations. From a clinical perspective, the significantly increased sample entropy compared to the normal ABPI group and the decreased and increased complex correlation properties measured by DFA for the low and high ABPI groups respectively, may be useful indicators that a more holistic treatment approach in line with this more complex clinical picture is required.

© 2015 Elsevier B.V. In many retinopathies one of the first alterations which can be detected is tortuosity of the retinal vessels. Already published techniques focused on the representation of the vessel as a mathematical curve, which provided good results but problems due to skeletonization or the choice of the sampling rate were shown recently. We present here a new algorithm for the automated grading of tortuosity, which we have applied to images created from the RET-TORT database. The algorithm is based on a combination of multiscale wavelet and nonlinear derived analysis and can be directly applied to images of segmented vessels without suffering from influences of imperfect mathematical abstractions or poorly chosen sampling rates. This improves reproducibility and is advantageous in clinical practice for identification of tortuosity and treatment decision making. Our method is robust against noise and provides equally good results for arterioles and venules, in line with manual rankings by specialists.

This work investigates the effectiveness of using computer-based machine learning regression algorithms and meta-regression methods to predict performance data for Australian football players based on parameters collected during daily physiological tests. Three experiments are described. The first uses all available data with a variety of regression techniques. The second uses a subset of features selected from the available data using the Random Forest method. The third used meta-regression with the selected feature subset. Our experiments demonstrate that feature selection and meta-regression methods improve the accuracy of predictions for match performance of Australian football players based on daily data of medical tests, compared to regression methods alone. Meta-regression methods and feature selection were able to obtain performance prediction outcomes with significant correlation coefficients. The best results were obtained by the additive regression based on isotonic regression for a set of most influential features selected by Random Forest. This model was able to predict athlete performance data with a correlation coefficient of 0.86 (p < 0.05). © 2013 Published by Elsevier B.V. All rights reserved.

Purpose ¿ This paper seeks to examine the spatial patterns of student use of machines in each laboratory to whether there are underlying commonalities. Design/methodology/approach ¿ The research was carried out by assessing the user behaviour in 16 computer laboratories at a regional university in Australia. Findings ¿ The study found that computers within easy access to doors are disproportionately more used than computer that are further away, irrespective of other ¿incentive¿ such as windows, wall anchoring or security camera positioning. Practical implications ¿ This paper has implications for any division within a university environment responsible for the spatial positioning of computer in a student laboratory. Originality/value ¿ Previous research of the use of computer laboratories in schools and universities has focussed on educational issues. None of the studies so far have considered matters of situational territoriality and spatial patterning that govern human behaviour. © 2007, Emerald Group Publishing Limited

Static resources allocation problems have been widely studied. More recently some of this attention has changed to focus on dynamic problems, where problem specifications, constraints or resources may change before a solution is obtained. This work examines an approach that combines a multi agent system based on a simulated market with evolutionary optimization. Previous work has showed the efficacy of such a hybrid approach, where the characteristics of agents are subject to evolutionary optimization. This work compares the multi agent only, and the hybrid system, when the problem is subject to random change during the attempt to find a solution. Results confirm the advantages of evolutionary optimization of agent rules in a static or dynamic environment, both in terms of tasks completed within a given time, and the cost per task completed. Surprisingly, an optimum amount of noise exists, that improves the performance of the multi agent or hybrid trading model. ©2007 IEEE.

Purpose - Many researchers require access to computer facilities beyond those offered by desktop workstations. Traditionally, these are offered either through partnerships, to share the cost of supercomputing facilities, or through purpose-built cluster facilities. However, funds are not always available to satisfy either of these options, and university departments are under increasing pressure to obtain the maximum return on investment. The purpose of this paper is to describe the development of a low-cost cluster computing facility using existing workstations in undergraduate computing laboratories. Design/methodology/approach - Drawing on a previous feasibility study, experiments were conducted with cluster configurations of increasing size to determine the benefits and drawbacks. This was followed by identification and resolution of problems, and assessment of effects upon existing users of the computers. Findings - A working cluster was built to supply the needs of researchers, with almost no impact upon existing users and at very little cost. Research limitations/implications - At present, the workstations can only be used as a "processor farm" and it is unclear how MPI type processing may be achieved, because there is no guarantee of the number of workstations available at any specific time. Practical implications - Any institution requiring access to high performance computing can build a useful resource from existing computers at minimal cost. The usefulness of the resource depends on existing computers being under utilised. Originality/value - The novel aspect of this work is that the facility was built using existing student computing laboratories. © Emerald Group Publishing Limited.

Purpose - To provide administrators at an Australian university with data on the feasibility of redirecting under-utilised computer laboratories facilities into a distributed high performance computing facility. Design/methodology/approach - The individual log-in records for each computer located in the computer laboratories at the university were investigated. The log-in data were investigated over a 24-hour/seven day a week period between June 2001 and August 2003. The data were analysed in terms of student access to the computer facilities during "normal" business hours, weekend times, and the semester breaks. Findings - The computer laboratories were hugely under-utilised, with less than 10 per cent of all log-ins occurring during off peak times (7 pm-8 am). Similarly, only weekends were likewise under-utilised. This strongly suggests that this spare computer capacity could be used for alternate means during these times. Research limitations/implications - Future research needs to determine whether the needs of the general computer laboratory user who requires a stable and secure system can coexist with the users of a high performance computer facility where different software and differently configured computer systems are required. Practical implications - This research has the potential for universities to utilise more effectively their computer laboratory resources by allocating under-utilised resources into other projects, such as to a high performance computing facility (HPCF). The cost of these re-allocated resources would be a fraction of the cost compared to a scenario in which a separate dedicated HPCF had to be provided. Originality/value - This paper suggests an alternate utilisation of the spare computing laboratory resources available at many universities. © Emerald Group Publishing Limited.

Automated image processing has the potential to assist in the early detection of diabetes, by detecting changes in blood vessel diameter and patterns in the retina. This paper describes the development of segmentation methodology in the processing of retinal blood vessel images obtained using non-mydriatic colour photography. The methods used include wavelet analysis, supervised classifier probabilities and adaptive threshold procedures, as well as morphology-based techniques. We show highly accurate identification of blood vessels for the purpose of studying changes in the vessel network that can be utilized for detecting blood vessel diameter changes associated with the pathophysiology of diabetes. In conjunction with suitable feature extraction and automated classification methods, our segmentation method could form the basis of a quick and accurate test for diabetic retinopathy, which would have huge benefits in terms of improved access to screening people for risk or presence of diabetes. © Copyright 2005.

Weasel World is an artificial life environment for simulating open ended evolution. It consists of a genetic algorithm embedded in a pseudo spatial context, composed of the text of Shakespeare's play Hamlet. The novelty of Weasel World is that it combines features from many evolutionary algorithms with an operation that increases genotype length, placement of individuals within a spatial context, the notion of territory, and interaction with global environmental parameters. Individuals interact with others in their territory, competing for the right to reproduce in a modified tournament selection. Weasel World demonstrates phase changes, clustering and self-organization, as well as showing the importance of considering the interaction between local phenomena and global constraints. The evolutionary process undergoes different phases arising from independent growth, a competitive phase and a lack of nutrient. © Copyright 2005.

© 2016 CCAL. Congestive Heart Failure (CHF) is a disease caused by the inability of the heart to supply the needs of the body in terms of oxygen and perfusion. Detection and diagnosis of CHF is difficult and requires a battery of tests, which include the electrocardiogram (ECG). Automated processing of the ECG signal and in particular heart rate variability (HRV) analysis holds great promise for diagnosis of CHF and more generally in assessing cardiac health, especially for personalized mobile health. However, recording the full 12-lead ECG is a relatively invasive procedure and for that reason it is of interest to determine what can be deduced from the much less intensive measurement of heart rate (RR interval) alone. In addition to calculating SDNN and, RMSSD, which when combined gave an accuracy of 78.8% with the Nearest Neighbour classifier. The best Renyi entropy result was an accuracy of 66.7% using Nearest Neighbour. Combining the best Renyi entropy results with SDNN and RMSSD led to an overall accuracy of 87.9% with sensitivity of 80% and specificity of 94.4%. In this work we have shown that applying Renyi entropy in addition to standard time domain measures identified CHF with higher accuracy than using time domain measures only. In addition, Renyi entropy exponents provide further information about the time signal characteristics that may be important in clinical decision making.

© Springer International Publishing Switzerland 2015. With the large and growing user base of social media, it is not an easy feat to identify potential customers for business. This is mainly due to the challenge of extracting commercially viable contents from the vast amount of free-form conversations. In this paper, we analyse the Twitter content of an account owner and its list of followers through various text mining methods and segment the list of followers via an index. We have termed this index as the High-Value Social Audience (HVSA) index. This HVSA index enables a company or organisation to devise their marketing and engagement plan according to available resources, so that a high-value social audience can potentially be transformed to customers, and hence improve the return on investment.

© 2015 IEEE. Heart rate variability (HRV) is commonly analysed in respect to time and frequency domain measures as well as nonlinear measures including the fractal dimension using information derived from ECG recordings. Increased risk of adverse cardiac events such as arrhythmia can be detected using heart rate variability analysis. Shannon entropy and the more generalized Renyi entropy provide information on the complexity of the interbeat interval differences associated with heart rate. The latter can be extended to a multiscale distribution akin to fractal analysis. Here we report the spectrum of multiscale Renyi entropy measures. Multiscale Renyi entropy has additional information to add to the common mean and variance measures and should be applied as potential early markers of arrhythmia risk.

Early identification of cardiac autonomic neuropathy (CAN) leads to better treatment outcomes. Heart rate variability (HRV) analysis allows identification of CAN but is sensitive to the length of recording and the presence of artifacts and ectopics, requiring preprocessing and consideration of length of recording. RR intervals from 10-second and 5-minute ECG recordings from patients with no CAN, early CAN and definite CAN were preprocessed using adaptive filtering with the controlling parameter c set at 0.2, 0.5 and 0.8, and time and frequency domain HRV analysis applied. Early CAN and definite CAN required different setting of c with respect to the length of recording. The 5-minute recording with c=0.2 provided the best results using RMSSD for normal versus eCAN (p=0.0007) and for eCAN versus dCAN (0.019). Clinically, this has potential use in general practice for screening patients at risk, newly diagnosed with diabetes, or for follow-up during the course of diabetes. © 2014 IEEE.

Early detection of cardiac autonomic neuropathy (CAN) is vital to provide timely health care. There are a variety of measures that can be considered for detection of CAN from the heart beat signal, which should provide a relatively noninvasive test. In this work we consider the use of two measures based on the RR interval - the standard deviation and the Renyi Entropy. We find that caution is indicated in the use of entropy for two reasons. First, these two measures are related, so that the researcher who provides Renyi entropy as a measure may in fact be measuring only the standard deviation. Second, there are several methods for calculating Renyi entropy, so the method needs to be carefully selected to provide meaningful results. © 2014 IEEE.

The Job-Shop Scheduling Problem (JSSP) is a well-known difficult combinatorial optimization problem. Many algorithms have been proposed for solving JSSP in the last few decades, including algorithms based on evolutionary techniques. However, there is room for improvement in solving medium to large scale problems effectively. In this paper, we present a Hybrid Genetic Algorithm (HGA) that includes a heuristic job ordering with a Genetic Algorithm. We apply HGA to a number of benchmark problems. It is found that the algorithm is able to improve the solution obtained by traditional genetic algorithm. © 2007 IEEE.

Many real world decision processes require to solve optimization problems. In this paper, an integrated Multiagent-Genetic Algorithm (MA-GA) is considered to solve constrained optimization problems. The applied approach is new in the literature for solving constrained optimization problems. Ten benchmark problems are used to test the performance of the approach and the results show impressive performance. © 2007 IEEE.

The Job-Shop Scheduling Problem (JSSP) is one of the most critical combinatorial optimization problems. The objective of JSSP in this research is to minimize the makespan. In this paper, we propose two Genetic Algorithm (GA) based approaches for solving JSSP. Firstly, we design a simple heuristic to reduce the completion time of jobs on the bottleneck machines that we call the reducing bottleneck technique (RBT). This heuristic was implemented in conjunction with a GA. Secondly; we propose to fill any possible gaps left in the simple GA solutions by the tasks that are scheduled later. We call this process the gap-utilization technique (GUT). With GUT, we also apply a swapping technique that deals only with the bottleneck job. We study 35 test problems with known solutions, using the existing GA and our proposed two algorithms. We obtain optimal solutions for 23 problems, and the solutions are very close for the rest. © 2007 IEEE.

In recent years, memetic algorithms (MAs) have been proposed to enhance the performance of evolutionary algorithms by incorporating local search techniques with evolutionary algorithms' global search ability, and applied successfully to solve different type of optimization problems. This paper proposes a new memetic algorithm and then introduces an agent-based memetic algorithm (AMA), for the first time, to further enhance the ability of MA in solving constrained optimization problems. In a lattice-like environment, each of the agents represents a candidate solution of the problem. The agents are able to sense and act on the society, and their performances i.e. fitness of the solution improves through co-evolutionary adaptation of society with the individual learning of the agents. The proposed algorithm is tested on 13 benchmark problems and the experimental results show promising performance. © 2007 IEEE.

During the last decade, both evolutionary computation and multiagent systems have been used for solving decision and optimization problems. This paper proposes a new evolutionary agent system by incorporating evolutionary process into agent concepts for solving mathematical programming models. Each of the agents represents a candidate solution of the problem, and able to sense and act on the society. The fitness of the agent improves through co-evolutionary adaptation of society with the individual learning of the agents. The performance of the proposed algorithm is tested on five new benchmark problems along with existing 13 well-known problems, and the experimental results show convincing performance. © Springer-Verlag Berlin Heidelberg 2007.

One focus of multi-agent systems research is the notion that complex outcomes or behaviours may be arrived at through the interaction of agents. However, it is still an open question as to how agents in a complex system form coalitions or modules, and how these coalitions self-organize into hierarchies. In this paper, we begin to address this question by investigating agent collaboration in the context of a high-level pattern recognition task. We propose a novel market-based communication protocol, which governs the aggregate behaviour of individual agents and subsequent emergent properties of the system. Based on the Contract Net Protocol, individual agents bid to join coalitions (or solutions to a given problem). An important contribution of this study is the analysis of the role heterogeneous agents play in the formation of coalitions. Using a simple model, we show that by promoting diversity within the agent population it is possible to avoid deadlock or "tie" conditions, which otherwise have to be solved arbitrarily by the deadlocked agents.

© Springer-Verlag Berlin Heidelberg 2002. Like many diseases, dementia is associated with a changed physical structure of diseased tissue. This study is a preliminary attempt to show that these changes are detectable using image processing, and could facilitate the automated classification of dementia subtypes. The identification of a link between different pathologies and the physical structure of tissue is potentially of great benefit to our understanding of this group of diseases. We have shown the existence of such a link by applying machine learning techniques to features derived using fractal analysis, as well as classical shape parameters.