The Combination of Multiple Classifiers Using an Evidential Reasoning Approach

In many domains when we have several competing classifiers available we want to synthesize them or some of them to get a more accurate classifier by a combination function. In this paper we propose a ‘class-indifferent’ method for combining classifier decisions represented by evidential structures called triplet and quartet, using Dempster's rule of combination. This method is unique in that it distinguishes important elements from the trivial ones in representing classifier decisions, makes use of more information than others in calculating the support for class labels and provides a practical way to apply the theoretically appealing Dempster–Shafer theory of evidence to the problem of ensemble learning. We present a formalism for modelling classifier decisions as triplet mass functions and we establish a range of formulae for combining these mass functions in order to arrive at a consensus decision. In addition we carry out a comparative study with the alternatives of simplet and dichotomous structure and also compare two combination methods, Dempster's rule and majority voting, over the UCI benchmark data, to demonstrate the advantage our approach offers. (A continuation of the work in this area that was published in IEEE Trans on KDE, and conferences)

Role of combination vibrations in resonant positron annihilation

Positrons can attach to molecules via vibrational Feshbach resonances, leading to very large annihilation rates. The predictions of a recent theory for this process are validated for deuterated methyl halides where all modes are dipole coupled to the incident positron. Data and analysis are presented for methanol and ethylene, demonstrating the importance of combination and overtone resonances and the ability of the theory to account for these features. The mechanism for these resonances and criteria for their occurrence as well as outstanding questions are discussed.

Robust face recognition with partial occlusion, illumination variation and limited training data by optimal feature selection

This study investigates face recognition with partial occlusion, illumination variation and their combination, assuming no prior information about the mismatch, and limited training data for each person. The authors extend their previous posterior union model (PUM) to give a new method capable of dealing with all these problems. PUM is an approach for selecting the optimal local image features for recognition to improve robustness to partial occlusion. The extension is in two stages. First, authors extend PUM from a probability-based formulation to a similarity-based formulation, so that it operates with as little as one single training sample to offer robustness to partial occlusion. Second, they extend this new formulation to make it robust to illumination variation, and to combined illumination variation and partial occlusion, by a novel combination of multicondition relighting and optimal feature selection. To evaluate the new methods, a number of databases with various simulated and realistic occlusion/illumination mismatches have been used. The results have demonstrated the improved robustness of the new methods.

The impact of nonadherence to inhaled long-acting β - adrenoceptor agonist/corticosteroid combination therapy on healthcare costs in difficult-to-control asthma

Background: Asthma is a leading, preventable cause of morbidity, mortality and cost. A disproportionate amount of the cost is generated by the 5-10%of patients with difficult-to-control asthma, who are prescribed treatment at step 4/5 of the Global Initiative for Asthma (GINA) guidelines. We have previously demonstrated a high prevalence of nonadherence to inhaled combination therapy (i.e. long-acting ß -adrenoceptor agonist [ß - agonist] and corticosteroid) in this population. The aim of this study was to examine the costs of healthcare utilization in a nonadherent group of patients with difficult-to-control asthma compared with adherent subjects. We also wished to examine potential savings if nonadherence to inhaled combination therapy could be addressed. All costs were measured from the perspective of a publicly funded health service Methods: Adherence was determined through examination of patient prescription refill behaviour and validated with a medical concordance interview. Data on healthcare use were collected from a patient survey and hospital records that included prescribed medicines, hospital admissions, intensive care unit (ICU) admissions and other unscheduled healthcare visits associated with asthma care. Activity was monetized using standard UK references and between-group comparisons based on a series of univariate and multivariate regression analyses. Results: Cost differences were identified for inhaled combination therapy, nebulizer, short acting b2-agonists and hospital costs excluding and including ICU admissions between adherent and nonadherent subjects. Compared with a group who have refractory asthma and who are adherent with medication, additional healthcare costs in nonadherent subjects are offset by the reduction in costs associated with reduced medication utilization. However, if nonadherence can be successfully targeted and hospital admissions avoided in this population, there is a potential $475 ($843-$368) saving per patient, per annum. Conclusion: Nonadherence is an important cause of difficult-to-control asthma. A uniform cost for subjects with difficult-to-control disease can be applied to economic analyses, independent of adherence, as increased healthcare utilization costs are offset by the reduced medication cost due to poor adherence. However, there are substantial potential savings in subjects with difficult-to-control asthma, who are nonadherent to inhaled combination therapy, if cost effective strategies for nonadherence are developed. © 2011 Adis Data Information BV. All rights reserved.

Resistance Development of Cystic Fibrosis Respiratory Pathogens When Exposed to Fosfomycin and Tobramycin Alone and in Combination under Aerobic and Anaerobic Conditions

Although antibiotics from different classes are frequently prescribed in combination to prevent the development of resistance amongst Cystic Fibrosis (CF) respiratory pathogens, there is a lack of data as to the efficacy of this approach. We have previously shown that a 4:1 (w/w) combination of fosfomycin and tobramycin (F:T) has excellent activity against CF pathogens with increased activity under physiologically relevant anaerobic conditions. Therefore, the aim of this study was to determine whether F:T could delay or prevent the onset of resistance compared to either fosfomycin or tobramycin alone under aerobic and anaerobic conditions. The frequency of spontaneous mutants arising following exposure to fosfomycin, tobramycin and F:T was determined for clinical Pseudomonas aeruginosa and MRSA isolates under aerobic and anaerobic conditions. The effect of sub-inhibitory concentrations of fosfomycin, tobramycin and F:T on the induction of resistance was also investigated, with the stability of resistance and fitness cost associated with resistance assessed if it developed. P. aeruginosa and MRSA isolates had a lower frequency of spontaneous mutants to F:T compared to fosfomycin and tobramycin under both aerobic and anaerobic conditions. There was a maximum two-fold increase in F:T MICs when P. aeruginosa and MRSA isolates were passaged in sub-inhibitory F:T for 12 days. In contrast, sequential resistance to fosfomycin and tobramycin developed quickly (n = 3 days for both) after passage in sub-inhibitory concentrations. Once developed, both fosfomycin and tobramycin resistance was stable and not associated with a biological fitness cost to either P. aeruginosa or MRSA isolates. The results of this study suggest that F:T may prevent the development of resistance compared to fosfomycin or tobramycin alone under aerobic and physiologically relevant anaerobic conditions. F:T may be a potential treatment option in CF patients chronically colonised by MRSA and/or P. aeruginosa.

Combination of CFD and DOE to analyze and improve the mass flow rate in urinary catheters

The urinary catheter is a thin plastic tube that has been designed to empty the bladder artificially, effortlessly, and with minimum discomfort. The current CH14 male catheter design was examined with a view to optimizing the mass flow rate. The literature imposed constraints to the analysis of the urinary catheter to ensure that a compromise between optimal flow, patient comfort, and everyday practicality from manufacture to use was achieved in the new design. As a result a total of six design characteristics were examined. The input variables in question were the length and width of eyelets 1 and 2 (four variables), the distance between the eyelets, and the angle of rotation between the eyelets. Due to the high number of possible input combinations a structured approach to the analysis of data was necessary. A combination of computational fluid dynamics (CFD) and design of experiments (DOE) has been used to evaluate the "optimal configuration." The use of CFD couple with DOE is a novel concept, which harnesses the computational power of CFD in the most efficient manner for prediction of the mass flow rate in the catheter. Copyright © 2009 by ASME.

Adaptive linear combination of heuristic orderings in constructing examination timetables

In this paper, we investigate adaptive linear combinations of graph coloring heuristics with a heuristic modifier to address the examination timetabling problem. We invoke a normalisation strategy for each parameter in order to generalise the specific problem data. Two graph coloring heuristics were used in this study (largest degree and saturation degree). A score for the difficulty of assigning each examination was obtained from an adaptive linear combination of these two heuristics and examinations in the list were ordered based on this value. The examinations with the score value representing the higher difficulty were chosen for scheduling based on two strategies. We tested for single and multiple heuristics with and without a heuristic modifier with different combinations of weight values for each parameter on the Toronto and ITC2007 benchmark data sets. We observed that the combination of multiple heuristics with a heuristic modifier offers an effective way to obtain good solution quality. Experimental results demonstrate that our approach delivers promising results. We conclude that this adaptive linear combination of heuristics is a highly effective method and simple to implement.

A randomised, phase II trial of the DNA-hypomethylating agent 5-aza-2'-deoxycytidine (decitabine) in combination with carboplatin vs carboplatin alone in patients with recurrent, partially platinum-sensitive ovarian cancer.

Background: Our previous laboratory and clinical data suggested that one mechanism underlying the development of platinum resistance in ovarian cancer is the acquisition of DNA methylation. We therefore tested the hypothesis that the DNA hypomethylating agent 5-aza-2'-deoxycytodine (decitabine) can reverse resistance to carboplatin in women with relapsed ovarian cancer.

Methods: Patients progressing 6-12 months after previous platinum therapy were randomised to decitabine on day 1 and carboplatin (AUC 6) on day 8, every 28 days or carboplatin alone. The primary objective was response rate in patients with methylated hMLH1 tumour DNA in plasma.

Results: After a pre-defined interim analysis, the study closed due to lack of efficacy and poor treatment deliverability in 15 patients treated with the combination. Responses by GCIG criteria were 9 out of 14 vs 3 out of 15 and by RECIST were 6 out of 13 vs 1 out of 12 for carboplatin and carboplatin/decitabine, respectively. Grade 3/4 neutropenia was more common with the combination (60% vs 15.4%) as was G2/3 carboplatin hypersensitivity (47% vs 21%).

Conclusions: With this schedule, the addition of decitabine appears to reduce rather than increase the efficacy of carboplatin in partially platinum-sensitive ovarian cancer and is difficult to deliver. Patient-selection strategies, different schedules and other demethylating agents should be considered in future combination studies.

Job strain and the risk of severe asthma exacerbations: A meta-analysis of individual-participant data from 100 000 European men and women

Background: Many patients and healthcare professionals believe that work-related psychosocial stress, such as job strain, can make asthma worse, but this is not corroborated by empirical evidence. We investigated the associations between job strain and the incidence of severe asthma exacerbations in working-age European men and women. Methods: We analysed individual-level data, collected between 1985 and 2010, from 102 175 working-age men and women in 11 prospective European studies. Job strain (a combination of high demands and low control at work) was self-reported at baseline. Incident severe asthma exacerbations were ascertained from national hospitalization and death registries. Associations between job strain and asthma exacerbations were modelled using Cox regression and the study-specific findings combined using random-effects meta-analyses. Results: During a median follow-up of 10 years, 1 109 individuals experienced a severe asthma exacerbation (430 with asthma as the primary diagnostic code). In the age- and sex-adjusted analyses, job strain was associated with an increased risk of severe asthma exacerbations defined using the primary diagnostic code (hazard ratio, HR: 1.27, 95% confidence interval, CI: 1.00, 1.61). This association attenuated towards the null after adjustment for potential confounders (HR: 1.22, 95% CI: 0.96, 1.55). No association was observed in the analyses with asthma defined using any diagnostic code (HR: 1.01, 95% CI: 0.86, 1.19). Conclusions: Our findings suggest that job strain is probably not an important risk factor for severe asthma exacerbations leading to hospitalization or death.

Thermal properties and eutectic behaviour of dapivirine in combination with steroid hormones and other antiretrovirals

Background: Combination drug products can display thermal behaviour that is more complex than for the corresponding single drug products. For example, the contraceptive vaginal ring (VR) Nuvaring contains a eutectic (lowest melting) composition of etonogestrel (ETN) and ethinyl estradiol. Here we report the predisposition of dapivirine (DPV) to form reduced melting/eutectic mixtures when combined with other contraceptive hormones and antiretrovirals, and discuss the implications for development of combination microbicide and multipurpose prevention technology (MPT) products.
Methods: Binary mixtures of DPV with darunavir (DRV), levonorgestrel (LNG), ETN or maraviroc (MVC) were prepared either by physical mixing or by solvent evaporation. Selected binary mixtures were also incorporated into silicone elastomer (SE) VR devices. Thermal behavior of the mixtures was analyzed using differential scanning calorimetry (DSC) operating in standard heating ramp mode (10 °C/min). DSC data were used to construct two component phase diagrams for each binary system.
Results: Drug mixtures typically showed reduced melting transitions for both drug components, with clear evidence for a eutectic mixture at a well-defined intermediate composition. Eutectic temperatures and compositions for the various mixtures were: 40% DPV / 60% ETN - 170°C; 25% DPV / 75% MVC - 172°C; 65% DPV / 35% LNG - 192°C. In each case, the eutectic composition was also detected when the drug mixtures were incorporated into SE VRs. For the DPV/DRV system, the thermal behaviour is complicated by desolvation from the darunavir ethanolate polymorph.
Conclusions: When DPV is combined with small molecular weight hydrophobic drugs, the melting temperature for both drugs is typically reduced to a degree dependent on the composition of the mixture. At specified compositions, a low melting eutectic system results. The formation of eutectic behavior in binary drug systems needs to be carefully characterised in order to define product performance and drug release.

Identification of vegetable oil botanical speciation in refined vegetable oil blends using an innovative combination of chromatographic and spectroscopic techniques

European Regulation 1169/2011 requires producers of foods that contain refined vegetable oils to label the oil types. A novel rapid and staged methodology has been developed for the first time to identify common oil species in oil blends. The qualitative method consists of a combination of a Fourier Transform Infrared (FTIR) spectroscopy to profile the oils and fatty acid chromatographic analysis to confirm the composition of the oils when required. Calibration models and specific classification criteria were developed and all data were fused into a simple decision-making system. The single lab validation of the method demonstrated the very good performance (96% correct classification, 100% specificity, 4% false positive rate). Only a small fraction of the samples needed to be confirmed with the majority of oils identified rapidly using only the spectroscopic procedure. The results demonstrate the huge potential of the methodology for a wide range of oil authenticity work.

Evaluation of information extraction techniques to label extracted data from e-commerce web page

Automatically determining and assigning shared and meaningful text labels to data extracted from an e-Commerce web page is a challenging problem. An e-Commerce web page can display a list of data records, each of which can contain a combination of data items (e.g. product name and price) and explicit labels, which describe some of these data items. Recent advances in extraction techniques have made it much easier to precisely extract individual data items and labels from a web page, however, there are two open problems: 1. assigning an explicit label to a data item, and 2. determining labels for the remaining data items. Furthermore, improvements in the availability and coverage of vocabularies, especially in the context of e-Commerce web sites, means that we now have access to a bank of relevant, meaningful and shared labels which can be assigned to extracted data items. However, there is a need for a technique which will take as input a set of extracted data items and assign automatically to them the most relevant and meaningful labels from a shared vocabulary. We observe that the Information Extraction (IE) community has developed a great number of techniques which solve problems similar to our own. In this work-in-progress paper we propose our intention to theoretically and experimentally evaluate different IE techniques to ascertain which is most suitable to solve this problem.

Inappropriate prescribing of combination inhalers in Northern Ireland: retrospective cross-sectional cohort study of prescribing practice in primary care

BACKGROUND: Asthma management guidelines advocate a stepwise approach to asthma therapy, including the addition of a long-acting bronchodilator to inhaled steroid therapy at step 3. This is almost exclusively prescribed as inhaled combination therapy.

AIMS: To examine whether asthma prescribing practice for inhaled combination therapy (inhaled corticosteroid/long-acting β2-agonist (ICS/LABA)) in primary care in Northern Ireland is in line with national asthma management guidelines.

METHODS: Using data from the Northern Ireland Enhanced Prescribing Database, we examined initiation of ICS/LABA in subjects aged 5-35 years in 2010.

RESULTS: A total of 2,640 subjects (67%) had no inhaled corticosteroid monotherapy (ICS) in the study year or six months of the preceding year (lead-in period) and, extending this to a 12-month lead-in period, 52% had no prior ICS. 41% of first prescriptions for ICS/LABA were dispensed in January to March. Prior to ICS/LABA prescription, in the previous six months only 17% had a short-acting β2-agonist (SABA) dispensed, 5% received oral steroids, and 17% received an antibiotic.

CONCLUSIONS: ICS/LABA therapy was initiated in the majority of young subjects with asthma without prior inhaled steroid therapy. Most prescriptions were initiated in the January to March period. However, the prescribing of ICS/LABA did not appear to be driven by asthma symptoms (17% received SABA in the previous 6 months) or severe asthma exacerbation (only 5% received oral steroids). Significant reductions in ICS/LABA, with associated cost savings, would occur if the asthma prescribing guidelines were followed in primary care.

Geotechnical monitoring using a combination of LIDAR, real-time dGPS, and electrical tomography to assess geotechnical risk of a coastal landslide.

The Antrim Coast Road stretching from the seaport of Larne in the East of Northern Ireland has a well-deserved reputation for being one of the most spectacular roads in Europe (Day, 2006). However the problematic geology; Jurassic Lias Clay and Triassic Mudstone overlain by Cretaceous Limestone and Tertiary Basalt, and environmental variables result in frequent instances of slope instability manifested in both shallow debris flows and occasional massive rotational movements, creating a geotechnical risk to this highway. This paper describes how a variety of techniques are being used to both assess instability and monitor movement of these active slopes near one site at Straidkilly Point, Glenarm. An in-depth understanding of the geology was obtained via boreholes, resistivity surveys and laboratory testing. Environmental variables recorded by an on-site weather station were correlated with measured pore water pressure and soil moisture infiltration data. Terrestrial LiDAR (TLS), with surveys carried out on a bi-monthly basis allowed for the generation of Digital Elevation Models (DEMs) of difference, highlighting areas of recent movement, accumulation and depletion. Morphology parameters were generated from the DEMs and include slope, curvature and multiple measures of roughness. Changes in the structure of the slope coupled with morphological parameters were characterised and linked to progressive failures from the temporal monitoring. In addition to TLS monitoring, Aerial LiDAR datasets were used for the spatio-morphological characterisation of the slope on a macro scale. A Differential Global Positioning System (dGPS) was also deployed on site to provide a real-time warning system for gross movements, which were also correlated with environmental conditions. Frequent electrical resistivity tomography (ERT) surveys were also implemented to provide a better understanding of long-term changes in soil moisture and help to define the complex geology. The paper describes how the data obtained via a diverse range of methods has been combined to facilitate a more informed management regime of geotechnical risk by the Northern Ireland Roads Service.