Learning sparse latent representation and distance metric for image retrieval

The performance of image retrieval depends critically on the semantic representation and the distance function used to estimate the similarity of two images. A good representation should integrate multiple visual and textual (e.g., tag) features and offer a step closer to the true semantics of interest (e.g., concepts). As the distance function operates on the representation, they are interdependent, and thus should be addressed at the same time. We propose a probabilistic solution to learn both the representation from multiple feature types and modalities and the distance metric from data. The learning is regularised so that the learned representation and information-theoretic metric will (i) preserve the regularities of the visual/textual spaces, (ii) enhance structured sparsity, (iii) encourage small intra-concept distances, and (iv) keep inter-concept images separated. We demonstrate the capacity of our method on the NUS-WIDE data. For the well-studied 13 animal subset, our method outperforms state-of-the-art rivals. On the subset of single-concept images, we gain 79:5% improvement over the standard nearest neighbours approach on the MAP score, and 45.7% on the NDCG.

Estimating the risk of cardiovascular disease using an obese-years metric

OBJECTIVE: To examine the association between obese-years and the risk of cardiovascular disease (CVD).

STUDY DESIGN: Prospective cohort study.

SETTING: Boston, USA.

PARTICIPANTS: 5036 participants of the Framingham Heart Study were examined.

METHODS: Obese-years was calculated by multiplying for each participant the number of body mass index (BMI) units above 29 kg/m(2) by the number of years lived at that BMI during approximately 50 years of follow-up. The association between obese-years and CVD was analysed using time-dependent Cox regression adjusted for potential confounders and compared with other models using the Akaike information criterion (AIC). The lowest AIC indicated better fit.

PRIMARY OUTCOME CVD RESULTS: The median cumulative obese-years was 24 (range 2-556 obese-years). During 138,918 person-years of follow-up, 2753 (55%) participants were diagnosed with CVD. The incidence rates and adjusted HR (AHR) for CVD increased with an increase in the number of obese-years. AHR for the categories 1-24.9, 25-49.9, 50-74.9 and ≥75 obese-years were, respectively, 1.31 (95% CI 1.15 to 1.48), 1.37 (95% CI 1.14 to 1.65), 1.62 (95% CI 1.32 to 1.99) and 1.80 (95% CI 1.54 to 2.10) compared with those who were never obese (ie, had zero obese-years). The effect of obese-years was stronger in males than females. For every 10 unit increase in obese-years, the AHR of CVD increased by 6% (95% CI 4% to 8%) for males and 3% (95% CI 2% to 4%) for females. The AIC was lowest for the model containing obese-years compared with models containing either the level of BMI or the duration of obesity alone.

CONCLUSIONS: This study demonstrates that obese-years metric conceptually captures the cumulative damage of obesity on body systems, and is found to provide slightly more precise estimation of the risk of CVD than the level or duration of obesity alone.

A new Raman metric for the characterisation of graphene oxide and its derivatives

Raman spectroscopy is among the primary techniques for the characterisation of graphene materials, as it provides insights into the quality of measured graphenes including their structure and conductivity as well as the presence of dopants. However, our ability to draw conclusions based on such spectra is limited by a lack of understanding regarding the origins of the peaks. Consequently, traditional characterisation techniques, which estimate the quality of the graphene material using the intensity ratio between the D and the G peaks, are unreliable for both GO and rGO. Herein we reanalyse the Raman spectra of graphenes and show that traditional methods rely upon an apparent G peak which is in fact a superposition of the G and D' peaks. We use this understanding to develop a new Raman characterisation method for graphenes that considers the D' peak by using its overtone the 2D'. We demonstrate the superiority and consistency of this method for calculating the oxygen content of graphenes, and use the relationship between the D' peak and graphene quality to define three regimes. This has important implications for purification techniques because, once GO is reduced beyond a critical threshold, further reduction offers limited gain in conductivity.

Using Patient Health Questionnaire-9 item parameters of a common metric resulted in similar depression scores compared to independent item response theory model reestimation

Objectives: To investigate the validity of a common depression metric in independent samples. Study Design and Setting: We applied a common metrics approach based on item-response theory for measuring depression to four German-speaking samples that completed the Patient Health Questionnaire (PHQ-9). We compared the PHQ item parameters reported for this common metric to reestimated item parameters that derived from fitting a generalized partial credit model solely to the PHQ-9 items. We calibrated the new model on the same scale as the common metric using two approaches (estimation with shifted prior and StockingeLord linking). By fitting a mixed-effects model and using BlandeAltman plots, we investigated the agreement between latent depression scores resulting from the different estimation models. Results: We found different item parameters across samples and estimation methods. Although differences in latent depression scores between different estimation methods were statistically significant, these were clinically irrelevant. Conclusion: Our findings provide evidence that it is possible to estimate latent depression scores by using the item parameters from a common metric instead of reestimating and linking a model. The use of common metric parameters is simple, for example, using a Web application (http://www.common-metrics.org) and offers a long-term perspective to improve the comparability of patient-reported outcome measures.

Experience report: anomaly detection of cloud application operations using log and cloud metric correlation analysis

Failure of application operations is one of the maincauses of system-wide outages in cloud environments. Thisparticularly applies to DevOps operations, such as backup,redeployment, upgrade, customized scaling, and migration that areexposed to frequent interference from other concurrent operations,configuration changes, and resources failure. However, currentpractices fail to provide a reliable assurance of correct execution ofthese kinds of operations. In this paper, we present an approach toaddress this problem that adopts a regression-based analysistechnique to find the correlation between an operation’s activity logsand the operation activity’s effect on cloud resources. Thecorrelation model is then used to derive assertion specifications,which can be used for runtime verification of running operations andtheir impact on resources. We evaluated our proposed approach onAmazon EC2 with 22 rounds of rolling upgrade operations whileother types of operations were running and random faults wereinjected. Our experiment shows that our approach successfullymanaged to raise alarms for 115 random injected faults, with aprecision of 92.3%.

Floristic and structural characteristics of a coastal heathland exhibiting symptoms of Phytophthora cinnamomi infestation in the eastern Otway Ranges, Victoria

The floristics and structure of heathland vegetation exhibiting symptoms of Phytophthora cinnamomi Rands infestation was assessed at two sites in heathlands at Anglesea, Victoria. There were significant effects in both floristics and structure. Thirteen heathland species were significantly less abundant in diseased areas and 23 species were more abundant. Diseased (infested) vegetation, when compared with non-diseased areas, had less cover of Xanthorrhoea australis and shrub species and a greater cover of sedges, grasses and open ground. Structural differences were observed between heights 0 and 0.6 m, with a decline in cover recorded in diseased vegetation. Non-metric multidimensional scaling ordination of the floristic data showed a clear separation of diseased and non-diseased vegetation and that changes in floristic composition post-infestation were similar at both sites. Although there was some evidence of regeneration of X. australis, the recovery capacity of other susceptible species at Anglesea is unknown. The long-term consequences of loss of species and structure in the eastern Otways mean that the vegetation is unlikely to return to former status, especially if the pathogen continues to reinfect.

Methodological and conceptual issues in the search for a relationship between animal body-size distributions and benthic habitat architecture

Benthic ecologists have studied the distribution of animal body sizes because it is a form of ‘taxon-free’ classification that may be a useful metric for describing variation within and between ecological communities. In particular, the idea that the allometry of physiological and life-history traits may control species composition and relative abundances implies a functional link between body-size distributions and communities. The physical structure of aquatic habitats has often been cited as the mechanism by which habitat may determine body-size distributions in communities. However, further progress is hindered by a lack of theoretical clarity regarding the mechanisms that connect body size to the characteristics of ecological communities, leading to methods that may obscure interesting trends in body-size data. This review examines the methodological and conceptual issues hindering progress in the search for a relationship between animal body size and habitat architecture and suggests ways to resolve these issues. Problems are identified with current methods for the measurement of animal body size, the data and measures used to quantify body-size distributions and the methods used to identify patterns therein. Fundamentally, renewed emphasis on the mechanisms by which animal body sizes are influenced by habitat architecture is required to refine methodology and synthesise results from pattern-seeking and mechanistic studies.

Density based fuzzy c-means clustering of non-convex patterns

We propose a new technique to perform unsupervised data classification (clustering) based on density induced metric and non-smooth optimization. Our goal is to automatically recognize multidimensional clusters of non-convex shape. We present a modification of the fuzzy c-means algorithm, which uses the data induced metric, defined with the help of Delaunay triangulation. We detail computation of the distances in such a metric using graph algorithms. To find optimal positions of cluster prototypes we employ the discrete gradient method of non-smooth optimization. The new clustering method is capable to identify non-convex overlapped d-dimensional clusters.

The relationship between channel metrics, strategy and marketing performance

The role of marketing channels is to implement marketing strategy. The difficulty of channel strategy is compounded by the emergence of e-channels and the need to integrate e-channels into traditional or “bricks and mortar” channels (Rowley 2002). As a result, managing performance across a greater number of channels with diverse characteristics is more difficult.

Organization and marketing performance is to some degree a function of the quality of channel implementation and particularly channel performance measurement. The channels literature suggests a “channel performance metric paradox”. Approaches to channel performance metrics have been mutually orthogonal or even negatively correlated. (Jeuland & Shugan 1983; Lewis & Lambert 1991; Larson & Lusch 1992). This paradox implies that it is impossible for all channel performance metrics to be maximized simultaneously and tradeoffs exist.

This paper proposes a research model and propositions which extend previous research and attempts to reconcile this “channel performance metric paradox”. The model assumes that testing the relationship between the Miles and Snow strategy types and a comprehensive range of channel performance metrics may explain this paradox. Previous implementation performance research has focused more on the Porter strategies rather than the Miles and Snow strategy types.

Cool temperate rainforest Lichens of Victoria, Australia: floristics and distribution

Lichen floristics and distribution were examined in cool temperate rainforests of Victoria, Australia. Forty sites were sampled, ten from each of the Otways, Yarra Ranges, Baw Baw, and Errinundra regions. A total of 165 lichen species were recorded, representing 38 fam- ilies. Seventeen species are new for Victoria. Non-metric Multidimensional Scaling showed three lichen groups that corresponded with the rainforest types in which they occurred. A sub-group of lichen species were considered widespread and represented the core of Victoria's rainforest lichen flora.

A fuzzy logic based approach for software testing

How to provide cost-effective strategies for Software Testing has been one of the research focuses in Software Engineering for a long time. Many researchers in Software Engineering have addressed the effectiveness and quality metric of Software Testing, and many interesting results have been obtained. However, one issue of paramount importance in software testing — the intrinsic imprecise and uncertain relationships within testing metrics — is left unaddressed. To this end, a new quality and effectiveness measurement based on fuzzy logic is proposed. Related issues like the software quality features and fuzzy reasoning for test project similarity measurement are discussed, which can deal with quality and effectiveness consistency between different test projects. Experiments were conducted to verify the proposed measurement using real data from actual software testing projects. Experimental results show that the proposed fuzzy logic based metrics is effective and efficient to measure and evaluate the quality and effectiveness of test projects.

Dynamical systems for discovering protein complexes and functional modules from biological networks

Recent advances in high throughput experiments and annotations via published literature have provided a wealth of interaction maps of several biomolecular networks, including metabolic, protein-protein, and protein-DNA interaction networks. The architecture of these molecular networks reveals important principles of cellular organization and molecular functions. Analyzing such networks, i.e., discovering dense regions in the network, is an important way to identify protein complexes and functional modules. This task has been formulated as the problem of finding heavy subgraphs, the Heaviest k-Subgraph Problem (k-HSP), which itself is NPhard. However, any method based on the k-HSP requires the parameter k and an exact solution of k-HSP may still end up as a “spurious” heavy subgraph, thus reducing its practicability in analyzing large scale biological networks. We proposed a new formulation, called the rank-HSP, and two dynamical systems to approximate its results. In addition, a novel metric, called the Standard deviation and Mean Ratio (SMR), is proposed for use in “spurious” heavy subgraphs to automate the discovery by setting a fixed threshold. Empirical results on both the simulated graphs and biological networks have demonstrated the efficiency and effectiveness of our proposal.

Baseline assessment of metals and hydrocarbons in the sediments of Lake Mulwala, Australia

An assessment of hydrocarbon and metal/metalloids (arsenic, cadmium, copper, lead, mercury, selenium, zinc) contamination in sediments from Lake Mulwala, Australia, was undertaken. The objectives of the study were: (i) to determine the extent of contamination in the lake sediments, compared to Australian and international sediment quality guidelines, and (ii) to attempt to identify the contaminant sources to the lake. With the exception of a few samples containing elevated levels of arsenic and/or mercury, the levels of all contaminants in the sediment samples taken from the lake were below the 'lower trigger' of the Australian Sediment Quality Guidelines that would warrant further investigation. High molecular weight hydrocarbons (up to 700 mg kg−1) were found in most sediment samples. Non-metric statistical analysis indicated that the contaminant distribution was different in different parts of the lake, with the lowest concentrations generally found at the influent to the lake. No definitive source(s) of contamination could be identified for either metalloids or hydrocarbons.