Search Google Appliance

Publications

| 1975 | 1976 | 1977 | 1978 | 1979 | 1980 | 1981 | 1982 | 1983 | 1984 | 1985 | 1986 | 1987 | 1988 | 1989 | 1990 | 1991 | 1992 | 1993 | 1994 | 1995 | 1996 | 1997 | 1998 | 1999 | 2000 | 2001 | 2002 | 2003 | 2004 | 2005 | 2006 | 2007 | 2008 | 2009 | 2010 | 2011 | 2012 | 2013 | 2014 | 2015 | 2016 | 2017

Chapters

Adams, J., Sibbritt, D., Broom, A., Kroll, T., Prior, J., Dunston, R., Leung, B., Davidson, P. & Andrews, G. 2017, 'Traditional, complementary and integrative medicine as self-care in chronic illness' in Adams, J. & et al (eds), Public Health and Health Services Research in Traditional, Complementary and Integrative Medicine: International Perspectives, Imperial College Press, London.

Stoianoff, N.P., Cahill, A. & Wright, E.A. 2017, 'Indigenous knowledge: what are the issues?' in Stoianoff, N.P. (ed), Indigenous Knowledge Forum: Comparative Systems for Recognising and Protecting Indigenous Knowledge and Culture, LexisNexis, pp. 11-37.

Wright, E.A., Cahill, A. & Stoianoff, N.P. 2017, 'Australia and Indigenous traditional knowledge' in Stoianoff, N.P. (ed), Indigenous Knowledge Forum: Comparative Systems for Recognising and Protecting Indigenous Knowledge and Culture, LexisNexis, pp. 39-68.

Journal articles

Anderson, C. & Ryan, L.M. 2017, 'A Comparison of Spatio-Temporal Disease Mapping Approaches Including an Application to Ischaemic Heart Disease in New South Wales, Australia.', Int J Environ Res Public Health, vol. 14, no. 2.
View/Download from: UTS OPUS or Publisher's site
View description>>

The field of spatio-temporal modelling has witnessed a recent surge as a result of developments in computational power and increased data collection. These developments allow analysts to model the evolution of health outcomes in both space and time simultaneously. This paper models the trends in ischaemic heart disease (IHD) in New South Wales, Australia over an eight-year period between 2006 and 2013. A number of spatio-temporal models are considered, and we propose a novel method for determining the goodness-of-fit for these models by outlining a spatio-temporal extension of the Moran's I statistic. We identify an overall decrease in the rates of IHD, but note that the extent of this health improvement varies across the state. In particular, we identified a number of remote areas in the north and west of the state where the risk stayed constant or even increased slightly.

Chai, R., Ling, S.H., San, P.P., Naik, G., Nguyen, T.N., Tran, Y., Craig, A. & Nguyen, H.T. 2017, 'Improving EEG-based Driver Fatigue Classification using Sparse-Deep Belief Networks', Frontiers in Neuroscience, vol. 11, no. 103, pp. 1-14.
View/Download from: UTS OPUS or Publisher's site
View description>>

This paper presents an improvement of classification performance for electroencephalography (EEG)-based driver fatigue classification between fatigue and alert states with the data collected from 43 participants. The system employs autoregressive (AR) modeling as the features extraction algorithm, and sparse-deep belief networks (sparse-DBN) as the classification algorithm. Compared to other classifiers, sparse-DBN is a semi supervised learning method which combines unsupervised learning for modeling features in the pre-training layer and supervised learning for classification in the following layer. The sparsity in sparse-DBN is achieved with a regularization term that penalizes a deviation of the expected activation of hidden units from a fixed low-level prevents the network from overfitting and is able to learn low-level structures as well as high-level structures. For comparison, the artificial neural networks (ANN), Bayesian neural networks (BNN) and original deep belief networks (DBN) classifiers are used. The classification results show that using AR feature extractor and DBN classifiers, the classification performance achieves an improved classification performance with a of sensitivity of 90.8%, a specificity of 90.4%, an accuracy of 90.6% and an area under the receiver operating curve (AUROC) of 0.94 compared to ANN (sensitivity at 80.8%, specificity at 77.8%, accuracy at 79.3% with AUC-ROC of 0.83) and BNN classifiers (sensitivity at 84.3%, specificity at 83%, accuracy at 83.6% with AUROC of 0.87). Using the sparse-DBN classifier, the classification performance improved further with sensitivity of 93.9%, a specificity of 92.3% and an accuracy of 93.1% with AUROC of 0.96. Overall, the sparse-DBN classifier improved accuracy by 13.8%, 9.5% and 2.5% over ANN, BNN and DBN classifiers respectively.

Chai, R., Naik, G.R., Ling, S.H. & Nguyen, H.T. 2017, 'Hybrid brain–computer interface for biomedical cyber-physical system application using wireless embedded EEG systems', BioMedical Engineering OnLine, vol. 16, no. 5, pp. 1-23.
View/Download from: UTS OPUS or Publisher's site

Chiarella, C., He, X.Z., Shi, L. & Wei, L. 2017, 'A behavioural model of investor sentiment in limit order markets', Quantitative Finance, vol. 17, no. 1, pp. 71-86.
View/Download from: Publisher's site
View description>>

© 2016 Informa UK Limited, trading as Taylor & Francis GroupBy incorporating behavioural sentiment in a model of a limit order market, we show that behavioural sentiment not only helps to replicate most of the stylized facts in limit order markets simultaneously, but it also plays a unique role in explaining those stylized facts that cannot be explained by noise trading, such as fat tails in the return distribution, long memory in the trading volume, an increasing and non-linear relationship between trade imbalance and mid-price returns, as well as the diagonal effect, or event clustering, in order submission types. The results show that behavioural sentiment is an important driving force behind many of the well-documented stylized facts in limit order markets.

Ding, C. & Tao, D. 2017, 'Pose-invariant face recognition with homography-based normalization', Pattern Recognition, vol. 66, pp. 144-152.
View/Download from: Publisher's site
View description>>

© 2016 Elsevier LtdPose-invariant face recognition (PIFR) refers to the ability that recognizes face images with arbitrary pose variations. Among existing PIFR algorithms, pose normalization has been proved to be an effective approach which preserves texture fidelity, but usually depends on precise 3D face models or at high computational cost. In this paper, we propose an highly efficient PIFR algorithm that effectively handles the main challenges caused by pose variation. First, a dense grid of 3D facial landmarks are projected to each 2D face image, which enables feature extraction in an pose adaptive manner. Second, for the local patch around each landmark, an optimal warp is estimated based on homography to correct texture deformation caused by pose variations. The reconstructed frontal-view patches are then utilized for face recognition with traditional face descriptors. The homography-based normalization is highly efficient and the synthesized frontal face images are of high quality. Finally, we propose an effective approach for occlusion detection, which enables face recognition with visible patches only. Therefore, the proposed algorithm effectively handles the main challenges in PIFR. Experimental results on four popular face databases demonstrate that the propose approach performs well on both constrained and unconstrained environments.

Du, B., Wang, Z., Zhang, L., Zhang, L., Liu, W., Shen, J. & Tao, D. 2017, 'Exploring Representativeness and Informativeness for Active Learning', IEEE Transactions on Cybernetics, vol. 47, no. 1, pp. 14-26.
View/Download from: Publisher's site
View description>>

How can we find a general way to choose the most suitable samples for training a classifier? Even with very limited prior information? Active learning, which can be regarded as an iterative optimization procedure, plays a key role to construct a refined training set to improve the classification performance in a variety of applications, such as text analysis, image recognition, social network modeling, etc. Although combining representativeness and informativeness of samples has been proven promising for active sampling, state-of-the-art methods perform well under certain data structures. Then can we find a way to fuse the two active sampling criteria without any assumption on data? This paper proposes a general active learning framework that effectively fuses the two criteria. Inspired by a two-sample discrepancy problem, triple measures are elaborately designed to guarantee that the query samples not only possess the representativeness of the unlabeled data but also reveal the diversity of the labeled data. Any appropriate similarity measure can be employed to construct the triple measures. Meanwhile, an uncertain measure is leveraged to generate the informativeness criterion, which can be carried out in different ways. Rooted in this framework, a practical active learning algorithm is proposed, which exploits a radial basis function together with the estimated probabilities to construct the triple measures and a modified best-versus-second-best strategy to construct the uncertain measure, respectively. Experimental results on benchmark datasets demonstrate that our algorithm consistently achieves superior performance over the state-of-the-art active learning algorithms.

Du, B., Zhang, M., Zhang, L., Hu, R. & Tao, D. 2017, 'PLTD: Patch-Based Low-Rank Tensor Decomposition for Hyperspectral Images', IEEE Transactions on Multimedia, vol. 19, no. 1, pp. 67-79.
View/Download from: Publisher's site
View description>>

© 1999-2012 IEEE.Recent years has witnessed growing interest in hyperspectral image (HSI) processing. In practice, however, HSIs always suffer from huge data size and mass of redundant information, which hinder their application in many cases. HSI compression is a straightforward way of relieving these problems. However, most of the conventional image encoding algorithms mainly focus on the spatial dimensions, and they need not consider the redundancy in the spectral dimension. In this paper, we propose a novel HSI compression and reconstruction algorithm via patch-based low-rank tensor decomposition (PLTD). Instead of processing the HSI separately by spectral channel or by pixel, we represent each local patch of the HSI as a third-order tensor. Then, the similar tensor patches are grouped by clustering to form a fourth-order tensor per cluster. Since the grouped tensor is assumed to be redundant, each cluster can be approximately decomposed to a coefficient tensor and three dictionary matrices, which leads to a low-rank tensor representation of both the spatial and spectral modes. The reconstructed HSI can then be simply obtained by the product of the coefficient tensor and dictionary matrices per cluster. In this way, the proposed PLTD algorithm simultaneously removes the redundancy in both the spatial and spectral domains in a unified framework. The extensive experimental results on various public HSI datasets demonstrate that the proposed method outperforms the traditional image compression approaches and other tensor-based methods.

Ferguson, C., Inglis, S.C., Newton, P.J., Middleton, S., Macdonald, P.S. & Davidson, P.M. 2017, 'Barriers and enablers to adherence to anticoagulation in heart failure with atrial fibrillation: patient and provider perspectives'.
View/Download from: UTS OPUS or Publisher's site
View description>>

Aims & Objectives The purpose of this study was to elucidate the barriers and enablers to adherence to anticoagulation in individuals with chronic heart failure (CHF) with concomitant atrial fibrillation (AF) from the perspective of patients and providers. Background CHF and AF commonly coexist and are associated with increased stroke risk and mortality. Oral anticoagulation significantly reduces stroke risk and improves outcomes. Yet, in approximately 30% of cases anticoagulation is not commenced for a variety of reasons. Design Qualitative study using narrative inquiry. Methods Data from face to face individual interviews with patients and information retrieved from healthcare file note review documented the clinician perspective. This study is a synthesis of the two data sources, obtained during patient clinical assessments as part of the Atrial Fibrillation And Stroke Thromboprophylaxis in hEart failuRe (AFASTER) Study. Results Patient choice and preference were important factors in anticoagulation decisions, including treatment burden, unfavourable or intolerable side effects and patient refusal. Financial barriers included cost of travel, medication cost and reimbursement. Psychological factors included psychiatric illness, cognitive impairment and depression. Social barriers included homelessness and the absence of a caregiver or lack of caregiver assistance. Clinician reticence included fear of falls, frailty, age, fear of bleeding and the challenges of multi-morbidity. Facilitators to successful prescription and adherence were caregiver support, reminders and routine, self-testing and the use of technology. Conclusions Many barriers remain to high risk individuals being prescribed anticoagulation for stroke prevention. There are a number of enabling factors that facilitate prescription and optimize treatment adherence. Nurses should challenge these treatment barriers and seek enabling factors to optimise therapy. Relevance to clinical practice Nurs...

Gholizadeh, L., Ali Khan, S., Vahedi, F. & Davidson, P.M. 2017, 'Sensitivity and specificity of Urdu version of the PHQ-9 to screen depression in patients with coronary artery disease.', Contemp Nurse, pp. 1-7.
View/Download from: Publisher's site
View description>>

BACKGROUND: The Patient Health Questionnaire (PHQ-9) possesses many characteristics of a good screening tool and has the capacity to be used for screening depression in patients with coronary artery disease (CAD). AIM: To examine the psychometric properties and criterion validity of the PHQ-9 to screen and detect depression in patients with CAD in Pakistan. DESIGN: In this validation study, 150 patients with CAD completed the Urdu version of the PHQ-9. The major depressive episode module of the Mini International Neuropsychiatric Interview (MINI) was used as the gold standard. RESULTS: The Urdu version of the PHQ-9 revealed a good internal consistency with Cronbach's alpha of 0.83. Optimal sensitivity (76%) and specificity (76%) were achieved using the cut-off score of PHQ-9 ≥6, with area under the ROC curve of 0.86. CONCLUSION: The Urdu version of the PHQ-9 has acceptable psychometric properties to screen and detect major depression in patients with CAD.

Ghosh, S., Li, J., Cao, L. & Ramamohanarao, K. 2017, 'Septic shock prediction for ICU patients via coupled HMM walking on sequential contrast patterns.', J Biomed Inform, vol. 66, pp. 19-31.
View/Download from: UTS OPUS or Publisher's site
View description>>

BACKGROUND AND OBJECTIVE: Critical care patient events like sepsis or septic shock in intensive care units (ICUs) are dangerous complications which can cause multiple organ failures and eventual death. Preventive prediction of such events will allow clinicians to stage effective interventions for averting these critical complications. METHODS: It is widely understood that physiological conditions of patients on variables such as blood pressure and heart rate are suggestive to gradual changes over a certain period of time, prior to the occurrence of a septic shock. This work investigates the performance of a novel machine learning approach for the early prediction of septic shock. The approach combines highly informative sequential patterns extracted from multiple physiological variables and captures the interactions among these patterns via coupled hidden Markov models (CHMM). In particular, the patterns are extracted from three non-invasive waveform measurements: the mean arterial pressure levels, the heart rates and respiratory rates of septic shock patients from a large clinical ICU dataset called MIMIC-II. EVALUATION AND RESULTS: For baseline estimations, SVM and HMM models on the continuous time series data for the given patients, using MAP (mean arterial pressure), HR (heart rate), and RR (respiratory rate) are employed. Single channel patterns based HMM (SCP-HMM) and multi-channel patterns based coupled HMM (MCP-HMM) are compared against baseline models using 5-fold cross validation accuracies over multiple rounds. Particularly, the results of MCP-HMM are statistically significant having a p-value of 0.0014, in comparison to baseline models. Our experiments demonstrate a strong competitive accuracy in the prediction of septic shock, especially when the interactions between the multiple variables are coupled by the learning model. CONCLUSIONS: It can be concluded that the novelty of the approach, stems from the integration of sequence-based physiological pa...

Gong, C., Liu, T., Tang, Y., Yang, J., Yang, J. & Tao, D. 2017, 'A Regularization Approach for Instance-Based Superset Label Learning.', IEEE Trans Cybern.
View/Download from: Publisher's site
View description>>

Different from the traditional supervised learning in which each training example has only one explicit label, superset label learning (SLL) refers to the problem that a training example can be associated with a set of candidate labels, and only one of them is correct. Existing SLL methods are either regularization-based or instance-based, and the latter of which has achieved state-of-the-art performance. This is because the latest instance-based methods contain an explicit disambiguation operation that accurately picks up the groundtruth label of each training example from its ambiguous candidate labels. However, such disambiguation operation does not fully consider the mutually exclusive relationship among different candidate labels, so the disambiguated labels are usually generated in a nondiscriminative way, which is unfavorable for the instance-based methods to obtain satisfactory performance. To address this defect, we develop a novel regularization approach for instance-based superset label (RegISL) learning so that our instance-based method also inherits the good discriminative ability possessed by the regularization scheme. Specifically, we employ a graph to represent the training set, and require the examples that are adjacent on the graph to obtain similar labels. More importantly, a discrimination term is proposed to enlarge the gap of values between possible labels and unlikely labels for every training example. As a result, the intrinsic constraints among different candidate labels are deployed, and the disambiguated labels generated by RegISL are more discriminative and accurate than those output by existing instance-based algorithms. The experimental results on various tasks convincingly demonstrate the superiority of our RegISL to other typical SLL methods in terms of both training accuracy and test accuracy.

He, X. & Shi, L. 2017, 'Index Portfolio and Welfare Analysis Under Heterogeneous Beliefs', Journal of Banking and Finance, vol. 75, pp. 64-79.
View/Download from: Publisher's site

He, X. & Treich, N. 2017, 'Prediction market prices under risk aversion and heterogeneous beliefs', Journal of Mathematical Economics, vol. 70, pp. 105-114.
View/Download from: Publisher's site

He, X.Z., Lütkebohmert, E. & Xiao, Y. 2017, 'Rollover Risk and Credit Risk under Time-varying Margin', Quantitative Finance, vol. 17, no. 3, pp. 455-469.

Ho-Le, T.P., Center, J.R., Eisman, J.A., Nguyen, H.T. & Nguyen, T.V. 2017, 'Prediction of Bone Mineral Density and Fragility Fracture by Genetic Profiling.', Journal of Bone and Mineral Research, vol. 32, no. 2, pp. 285-293.
View/Download from: UTS OPUS or Publisher's site
View description>>

Although the susceptibility to fracture is partly determined by genetic factors, the contribution of newly discovered genetic variants to fracture prediction is still unclear. This study sought to define the predictive value of a genetic profiling for fracture prediction.Sixty-two bone mineral density (BMD)-associated single-nucleotide polymorphism (SNP) were genotyped in 557 men and 902 women who had participated in the Dubbo Osteoporosis Epidemiology Study. The incidence of fragility fracture was ascertained from X-ray reports between 1990 and 2015. Femoral neck BMD was measured by dual-energy X-ray absorptiometry. A weighted polygenic risk score (GRS) was created as a function of the number of risk alleles and their BMD-associated regression coefficients for each SNP. The association between GRS and fracture risk was assessed by the Cox's proportional hazards model.Individuals with greater GRS had lower femoral neck BMD (P < 0.01), but the variation in GRS accounted for less than 2% of total variance in BMD. Each unit increase in GRS was associated with a hazard ratio of 1.20 (95%CI, 1.04-1.38) for fracture, and this association was independent of age, prior fracture, fall, and in a subset of 33 SNPs, independent of femoral neck BMD. The significant association between GRS and fracture was observed for the vertebral and wrist fractures, but not for hip fracture. The area under the receiver operating characteristic (ROC) curve for the model with GRS and clinical risk factors was 0.71 (95% CI, 0.68-0.74). With GRS, the correct reclassification of fracture vs non-fracture ranged from 12% for hip fracture to 23% for wrist fracture.A genetic profiling of BMD-associated genetic variants could improve the accuracy of fracture prediction over and above that of clinical risk factors alone, and help stratify individuals by fracture status. This article is protected by copyright. All rights reserved.

Huang, T., Huang, M.L., Nguyen, Q., Zhao, L., Huang, W. & Chen, J. 2017, 'A Space-Filling Multidimensional Visualization (SFMDVis) for Exploratory Data Analysis', Information Sciences, vol. 390, pp. 32-53.
View/Download from: UTS OPUS or Publisher's site
View description>>

The space-filling visualization model was first invented by Ben Shneiderman [28] for maximizing the utilization of display space in relational data (or graph) visualization, especially for tree visualization. It uses the concept of Enclosure which dismisses the “edges” in the graphic representation that are all too frequently used in traditional node-link based graph visualizations. Therefore, the major issue in graph visualization which is the edge crossing can be naturally solved through the adoption of a space filling approach. However in the past, the space-filling concept has not attracted much attention from researchers in the field of multidimensional visualization. Although the problem of ‘edge crossing’ has also occurred among polylines which are used as the basic visual elements in the parallel coordinates visualization, it is problematic if those ‘edge crossings’ among polylines are not evenly distributed on the display plate as visual clutter will occur. This problem could significantly reduce the human readability in terms of reviewing a particular region of the visualization. In this study, we propose a new Space-Filling Multidimensional Data Visualization (SFMDVis) that for the first-time introduces a space-filling approach into multidimensional data visualization. The main contributions are: (1) achieving the maximization of space utilization in multidimensional visualization (i.e. 100% of the display area is fully used), (2) eliminating visual clutter in SFMDVis through the use of the non-classic geometric primitive and (3) improving the quality of visualization for the visual perception of linear correlations among different variables as well as recognizing data patterns. To evaluate the quality of SFMDVis, we have conducted a usability study to measure the performance of SFMDVis in comparison with parallel coordinates and a scatterplot matrix for finding linear correlations and data patterns. The evaluation results have suggested that the acc...

Liu, Q., Sun, Y., Wang, C., Liu, T. & Tao, D. 2017, 'Elastic net hypergraph learning for image clustering and semi-supervised classification', IEEE Transactions on Image Processing, vol. 26, no. 1, pp. 452-463.
View/Download from: Publisher's site
View description>>

© 1992-2012 IEEE.Graph model is emerging as a very effective tool for learning the complex structures and relationships hidden in data. In general, the critical purpose of graph-oriented learning algorithms is to construct an informative graph for image clustering and classification tasks. In addition to the classical K -nearest-neighbor and r-neighborhood methods for graph construction, l1-graph and its variants are emerging methods for finding the neighboring samples of a center datum, where the corresponding ingoing edge weights are simultaneously derived by the sparse reconstruction coefficients of the remaining samples. However, the pairwise links of l1-graph are not capable of capturing the high-order relationships between the center datum and its prominent data in sparse reconstruction. Meanwhile, from the perspective of variable selection, the l1 norm sparse constraint, regarded as a LASSO model, tends to select only one datum from a group of data that are highly correlated and ignore the others. To simultaneously cope with these drawbacks, we propose a new elastic net hypergraph learning model, which consists of two steps. In the first step, the robust matrix elastic net model is constructed to find the canonically related samples in a somewhat greedy way, achieving the grouping effect by adding the l2 penalty to the l1 constraint. In the second step, hypergraph is used to represent the high order relationships between each datum and its prominent samples by regarding them as a hyperedge. Subsequently, hypergraph Laplacian matrix is constructed for further analysis. New hypergraph learning algorithms, including unsupervised clustering and multi-class semi-supervised classification, are then derived. Extensive experiments on face and handwriting databases demonstrate the effectiveness of the proposed method.

Liu, T., Tao, D., Song, M. & Maybank, S. 2017, 'Algorithm-Dependent Generalization Bounds for Multi-Task Learning.', IEEE Trans Pattern Anal Mach Intell, vol. 39, no. 2, pp. 227-241.
View/Download from: Publisher's site
View description>>

Often, tasks are collected for multi-task learning (MTL) because they share similar feature structures. Based on this observation, in this paper, we present novel algorithm-dependent generalization bounds for MTL by exploiting the notion of algorithmic stability. We focus on the performance of one particular task and the average performance over multiple tasks by analyzing the generalization ability of a common parameter that is shared in MTL. When focusing on one particular task, with the help of a mild assumption on the feature structures, we interpret the function of the other tasks as a regularizer that produces a specific inductive bias. The algorithm for learning the common parameter, as well as the predictor, is thereby uniformly stable with respect to the domain of the particular task and has a generalization bound with a fast convergence rate of order O(1=n), where n is the sample size of the particular task. When focusing on the average performance over multiple tasks, we prove that a similar inductive bias exists under certain conditions on the feature structures. Thus, the corresponding algorithm for learning the common parameter is also uniformly stable with respect to the domains of the multiple tasks, and its generalization bound is of the order O(1=T ), where T is the number of tasks. These theoretical analyses naturally show that the similarity of feature structures in MTL will lead to specific regularizations for predicting, which enables the learning algorithms to generalize fast and correctly from a few examples.

Lv, L., Fan, S., Huang, M., Huang, W. & Yang, G. 2017, 'Golden Rectangle Treemap', Journal of Physics: Conference Series, vol. 787, no. 1, pp. 1-6.
View/Download from: UTS OPUS or Publisher's site
View description>>

Treemaps, a visualization method of representing hierarchical data sets, are becoming more and more popular for its efficient and compact displays. Several algorithms have been proposed to create more useful display by controlling the aspect ratios of the rectangles that make up a treemap. In this paper, we introduce a new treemap algorithm, generating layout in which the rectangles are easier to select and hierarchy information is easier to obtain. This algorithm generates rectangles which approximate golden rectangles. To prove the effectiveness of our algorithm, at the end of this paper several analyses on golden rectangle treemap have been done on disk file system.

Maneze, D., Ramjan, L., DiGiacomo, M., Everett, B., Davidson, P.M. & Salamonson, Y. 2017, 'Negotiating health and chronic illness in Filipino-Australians: A qualitative study with implications for health promotion', Ethnicity and Health.

Qiao, M., Liu, L., Yu, J., Xu, C. & Tao, D. 2017, 'Diversified dictionaries for multi-instance learning', Pattern Recognition, vol. 64, pp. 407-416.
View/Download from: Publisher's site
View description>>

© 2016 Elsevier LtdMultiple-instance learning (MIL) has been a popular topic in the study of pattern recognition for years due to its usefulness for such tasks as drug activity prediction and image/text classification. In a typical MIL setting, a bag contains a bag-level label and more than one instance/pattern. How to bridge instance-level representations to bag-level labels is a key step to achieve satisfactory classification accuracy results. In this paper, we present a supervised learning method, diversified dictionaries MIL, to address this problem. Our approach, on the one hand, exploits bag-level label information for training class-specific dictionaries. On the other hand, it introduces a diversity regularizer into the class-specific dictionaries to avoid ambiguity between them. To the best of our knowledge, this is the first time that the diversity prior is introduced to solve the MIL problems. Experiments conducted on several benchmark (drug activity and image/text annotation) datasets show that the proposed method compares favorably to state-of-the-art methods.

Rao, A., Newton, P., DiGiacomo, M., Hickman, L., Hwang, C. & Davidson, P. 2017, 'Optimal gender specific strategies for the secondary prevention of cardiovascular disease in women: a systematic review', Journal of Cardiopulmonary Rehabilitation and Prevention.

Rihari-Thomas, J., DiGiacomo, M., Phillips, J., Newton, P. & Davidson, P.M. 2017, 'Clinician Perspectives of Barriers to Effective Implementation of a Rapid Response System in an Academic Health Centre: A Focus Group Study', Int J Health Policy Manag, vol. 6, no. x, pp. 1-10.
View/Download from: UTS OPUS or Publisher's site
View description>>

Background: Systemic and structural issues of rapid response system (RRS) models can hinder implementation. This study sought to understand the ways in which acute care clinicians (physicians and nurses) experience and negotiate care for deteriorating patients within the RRS. Methods: Physicians and nurses working within an Australian academic health centre within a jurisdictional-based model of clinical governance participated in focus group interviews. Verbatim transcripts were analysed using thematic content analysis. Results: Thirty-four participants (21 physicians and 13 registered nurses [RNs]) participated in six focus groups over five weeks in 2014. Implementing the RRS in daily practice was a process of informal communication and negotiation in spite of standardised protocols. Themes highlighted several systems or organisational-level barriers to an effective RRS, including (1) responsibility is inversely proportional to clinical experience; (2) actions around system flexibility contribute to deviation from protocol; (3) misdistribution of resources leads to perceptions of inadequate staffing levels inhibiting full optimisation of the RRS; and (4) poor communication and documentation of RRS increases clinician workloads. Conclusion: Implementing a RRS is complex and multifactorial, influenced by various inter- and intra-professional factors, staffing models and organisational culture. The RRS is not a static model; it is both reflexive and iterative, perpetually transforming to meet healthcare consumer and provider demands and local unit contexts and needs. Requiring more than just a strong initial implementation phase, new models of care such as a RRS demand good governance processes, ongoing support and regular evaluation and refinement. Cultural, organizational and professional factors, as well as systems-based processes, require consideration if RRSs are to achieve their intended outcomes in dynamic healthcare settings.

Tian, D. & Tao, D. 2017, 'Global Hashing System for Fast Image Search', IEEE Transactions on Image Processing, vol. 26, no. 1, pp. 49-89.
View/Download from: Publisher's site
View description>>

© 1992-2012 IEEE.Hashing methods have been widely investigated for fast approximate nearest neighbor searching in large data sets. Most existing methods use binary vectors in lower dimensional spaces to represent data points that are usually real vectors of higher dimensionality. We divide the hashing process into two steps. Data points are first embedded in a low-dimensional space, and the global positioning system method is subsequently introduced but modified for binary embedding. We devise dataindependent and data-dependent methods to distribute the satellites at appropriate locations. Our methods are based on finding the tradeoff between the information losses in these two steps. Experiments show that our data-dependent method outperforms other methods in different-sized data sets from 100k to 10M. By incorporating the orthogonality of the code matrix, both our data-independent and data-dependent methods are particularly impressive in experiments on longer bits.

Walczak, A., Butow, P.N., Tattersall, M.H.N., Davidson, P.M., Young, J., Epstein, R.M., Costa, D.S.J. & Clayton, J.M. 2017, 'Encouraging early discussion of life expectancy and end-of-life care: A randomised controlled trial of a nurse-led communication support program for patients and caregivers', International Journal of Nursing Studies, vol. 67, pp. 31-40.
View/Download from: Publisher's site
View description>>

© 2016 Elsevier LtdBackground Patients are often not given the information needed to understand their prognosis and make informed treatment choices, with many consequently experiencing less than optimal care and quality-of-life at end-of-life. Objectives To evaluate the efficacy of a nurse-facilitated communication support program for patients with advanced, incurable cancer to assist them in discussing prognosis and end-of-life care. Design A parallel-group randomised controlled trial design was used. Settings This trial was conducted at six cancer treatment centres affiliated with major hospitals in Sydney, Australia. Participants 110 patients with advanced, incurable cancer participated. Methods The communication support program included guided exploration of a question prompt list, communication challenges, patient values and concerns and the value of discussing end-of-life care early, with oncologists cued to endorse question-asking and question prompt list use. Patients were randomised after baseline measure completion, a regular oncology consultation was audio-recorded and a follow-up questionnaire was completed one month later. Communication, health-related quality-of-life and satisfaction measures and a manualised consultation-coding scheme were used. Descriptive, Mixed Modelling and Generalised Linear Mixed Modelling analyses were conducted using SPSS version 22. Results Communication support program recipients gave significantly more cues for discussion of prognosis, end-of-life care, future care options and general issues not targeted by the intervention during recorded consultations, but did not ask more questions about these issues or overall. Oncologists’ question prompt list and question asking endorsement was inconsistent. Communication support program recipients’ self-efficacy in knowing what questions to ask their doctor significantly improved at follow-up while control arm patients’ self-efficacy declined. The communication support program did...

Wang, H., Zhang, P., Zhu, X., Tsang, I.W.H., Chen, L., Zhang, C. & Wu, X. 2017, 'Incremental Subgraph Feature Selection for Graph Classification', IEEE Transactions on Knowledge and Data Engineering, vol. 29, no. 1, pp. 128-142.
View/Download from: Publisher's site
View description>>

© 2016 IEEE.Graph classification is an important tool for analyzing data with structure dependency, where subgraphs are often used as features for learning. In reality, the dimension of the subgraphs crucially depends on the threshold setting of the frequency support parameter, and the number may become extremely large. As a result, subgraphs may be incrementally discovered to form a feature stream and require the underlying graph classifier to effectively discover representative subgraph features from the subgraph feature stream. In this paper, we propose a primal-dual incremental subgraph feature selection algorithm (ISF) based on a max-margin graph classifier. The ISF algorithm constructs a sequence of solutions that are both primal and dual feasible. Each primal-dual pair shrinks the dual gap and renders a better solution for the optimal subgraph feature set. To avoid bias of ISF algorithm on short-pattern subgraph features, we present a new incremental subgraph join feature selection algorithm (ISJF) by forcing graph classifiers to join short-pattern subgraphs and generate long-pattern subgraph features. We evaluate the performance of the proposed models on both synthetic networks and real-world social network data sets. Experimental results demonstrate the effectiveness of the proposed methods.

Xiong, W., Zhang, L., Du, B. & Tao, D. 2017, 'Combining local and global: Rich and robust feature pooling for visual recognition', Pattern Recognition, vol. 62, pp. 225-235.
View/Download from: Publisher's site
View description>>

© 2016 Elsevier LtdThe human visual system proves expert in discovering patterns in both global and local feature space. Can we design a similar way for unsupervised feature learning? In this paper, we propose a novel spatial pooling method within an unsupervised feature learning framework, named Rich and Robust Feature Pooling (R2FP), to better extract rich and robust representation from sparse feature maps learned from the raw data. Both local and global pooling strategies are further considered to instantiate such a method. The former selects the most representative features in the sub-region and summarizes the joint distribution of the selected features, while the latter is utilized to extract multiple resolutions of features and fuse the features with a feature balance kernel for rich representation. Extensive experiments on several image recognition tasks demonstrate the superiority of the proposed method.

Xu, Z., Tao, D., Huang, S. & Zhang, Y. 2017, 'Friend or Foe: Fine-Grained Categorization with Weak Supervision', IEEE Transactions on Image Processing, vol. 26, no. 1, pp. 135-146.
View/Download from: Publisher's site
View description>>

© 2016 IEEE.Multi-instance learning (MIL) is widely acknowledged as a fundamental method to solve weakly supervised problems. While MIL is usually effective in standard weakly supervised object recognition tasks, in this paper, we investigate the applicability of MIL on an extreme case of weakly supervised learning on the task of fine-grained visual categorization, in which intra-class variance could be larger than inter-class due to the subtle differences between subordinate categories. For this challenging task, we propose a new method that generalizes the standard multi-instance learning framework, for which a novel multi-task co-localization algorithm is proposed to take advantage of the relationship among fine-grained categories and meanwhile performs as an effective initialization strategy for the non-convex multi-instance objective. The localization results also enable object-level domain-specific fine-tuning of deep neural networks, which significantly boosts the performance. Experimental results on three fine-grained datasets reveal the effectiveness of the proposed method, especially the importance of exploiting inter-class relationships between object categories in weakly supervised fine-grained recognition.

Zhang, T., Jia, W., Yang, B., Yang, J., He, X. & Zheng, Z. 2017, 'MoWLD: a robust motion image descriptor for violence detection', Multimedia Tools and Applications, vol. 76, no. 1, pp. 1419-1438.
View/Download from: UTS OPUS or Publisher's site
View description>>

© 2015 Springer Science+Business Media New York Automatic violence detection from video is a hot topic for many video surveillance applications. However, there has been little success in designing an algorithm that can detect violence in surveillance videos with high performance. Existing methods typically apply the Bag-of-Words (BoW) model on local spatiotemporal descriptors. However, traditional spatiotemporal features are not discriminative enough, and also the BoW model roughly assigns each feature vector to only one visual word and therefore ignores the spatial relationships among the features. To tackle these problems, in this paper we propose a novel Motion Weber Local Descriptor (MoWLD) in the spirit of the well-known WLD and make it a powerful and robust descriptor for motion images. We extend the WLD spatial descriptions by adding a temporal component to the appearance descriptor, which implicitly captures local motion information as well as low-level image appear information. To eliminate redundant and irrelevant features, the non-parametric Kernel Density Estimation (KDE) is employed on the MoWLD descriptor. In order to obtain more discriminative features, we adopt the sparse coding and max pooling scheme to further process the selected MoWLDs. Experimental results on three benchmark datasets have demonstrated the superiority of the proposed approach over the state-of-the-arts.

Conferences

Chinchore, A., Xu, G. & Jiang, F. 2017, 'Classifying sybil in MSNs using C4.5', IEEE/ACM BESC 2016 - Proceedings of 2016 International Conference on Behavioral, Economic, Socio - Cultural Computing.
View/Download from: Publisher's site
View description>>

© 2016 IEEE.Sybil detection is an important task in cyber security research. Over past years, many data mining algorithms have been adopted to fulfill such task. Using classification and regression for sybil detection is a very challenging task. Despite of existing research made toward modeling classification for sybil detection and prediction, this research has proposed new solution on how sybil activity could be tracked to address this challenging issue. Prediction of sybil behaviour has been demonstrated by analysing the graph-based classification and regression techniques, using decision trees and described dependencies across different methods. Calculated gain and maxGain helped to trace some sybil users in the datasets.

Jiang, F., Gan, J., Xu, Y. & Xu, G. 2017, 'Coupled behavioral analysis for user preference-based email spamming', IEEE/ACM BESC 2016 - Proceedings of 2016 International Conference on Behavioral, Economic, Socio - Cultural Computing.
View/Download from: Publisher's site
View description>>

© 2016 IEEE.In this paper, we develop and implement a new email spamming system leveraged by coupled text similarity analysis on user preference and a virtual meta-layer user-based email network, we take the social networks or campus LAN networks as the spam social network scenario. Fewer current practices exploit social networking initiatives to assist in spam filtering. Social network has essentially a large number of accounts features and attributes to be considered. Instead of considering large amount of users accounts features, we construct a new model called meta-layer email network which can reduce these features by only considering individual user's actions as an indicator of user preference, these common user actions are considered to construct a social behavior-based email network. With the further analytic results from text similarity measurements for each individual email contents, the behavior-based virtual email network can be improved with much higher accuracy on user preferences. Further, a coupled selection model is developed for this email network, we are able to consider all relevant factors/features in a whole and recommend the emails practically to the user individually. The experimental results show the new approach can achieve higher precision and accuracy with better email ranking in favor of personalised preference.

Li, Y., Tian, X. & Tao, D. 2017, 'Regularized large margin distance metric learning', Proceedings - IEEE International Conference on Data Mining, ICDM, pp. 1015-1022.
View/Download from: Publisher's site
View description>>

© 2016 IEEE.Distance metric learning plays an important role in many applications, such as classification and clustering. In this paper, we propose a novel distance metric learning using two hinge losses in the objective function. One is the constraint of the pairs which makes the similar pairs (the same label) closer and the dissimilar (different labels) pairs separated as far as possible. The other one is the constraint of the triplets which makes the largest distance between pairs intra the class larger than the smallest distance between pairs inter the classes. Previous works only consider one of the two kinds of constraints. Additionally, different from the triplets used in previous works, we just need a small amount of such special triplets. This improves the efficiency of our proposed method. Consider the situation in which we might not have enough labeled samples, we extend the proposed distance metric learning into a semi-supervised learning framework. Experiments are conducted on several landmark datasets and the results demonstrate the effectiveness of our proposed method.

Other

Aliyev, N. & He, X. 2017, 'Ambiguous market making', SSRN.

He, X., Li, K. & Shi, L. 2017, 'Social interactions, stochastic volatility, and momentum', SSRN.