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)

Fusion of infrared polarization and intensity images using support value transform and fuzzy combination rules

Infrared polarization and intensity imagery provide complementary and discriminative information in image understanding and interpretation. In this paper, a novel fusion method is proposed by effectively merging the information with various combination rules. It makes use of both low-frequency and highfrequency images components from support value transform (SVT), and applies fuzzy logic in the combination process. Images (both infrared polarization and intensity images) to be fused are firstly decomposed into low-frequency component images and support value image sequences by the SVT. Then the low-frequency component images are combined using a fuzzy combination rule blending three sub-combination methods of (1) region feature maximum, (2) region feature weighting average, and (3) pixel value maximum; and the support value image sequences are merged using a fuzzy combination rule fusing two sub-combination methods of (1) pixel energy maximum and (2) region feature weighting. With the variables of two newly defined features, i.e. the low-frequency difference feature for low-frequency component images and the support-value difference feature for support value image sequences, trapezoidal membership functions are proposed and developed in tuning the fuzzy fusion process. Finally the fused image is obtained by inverse SVT operations. Experimental results of visual inspection and quantitative evaluation both indicate the superiority of the proposed method to its counterparts in image fusion of infrared polarization and intensity images.

Effect of Three Methods for Cleaning Dentures on Biofilms Formed In Vitro on Acrylic Resin

Purpose: The aim of this study was to evaluate the effect of three denture hygiene methods against different microbial biofilms formed on acrylic resin specimens. Materials and methods: The set (sterile stainless steel basket and specimens) was contaminated (37 degrees C for 48 hours) by a microbial inoculum with 106 colony-forming units (CFU)/ml (standard strains: Staphylococcus aureus, Streptococcus mutans, Escherichia coli, Candida albicans, Pseudomonas aeruginosa, and Enterococcus faecalis; field strains: S. mutans, C. albicans, C. glabrata, and C. tropicalis). After inoculation, specimens were cleansed by the following methods: (1) chemical: immersion in an alkaline peroxide solution (Bonyplus tablets) for 5 minutes; (2) mechanical: brushing with a dentifrice for removable prostheses (Dentu Creme) for 20 seconds; and (3) a combination of chemical and mechanical methods. Specimens were applied onto a Petri plate with appropriate culture medium for 10 minutes. Afterward, the specimens were removed and the plates incubated at 37 degrees C for 48 hours. Results: Chemical, mechanical, and combination methods showed no significant difference in the reduction of CFU for S. aureus, S. mutans (ATCC and field strain), and P. aeruginosa. Mechanical and combination methods were similar and more effective than the chemical method for E. faecalis, C. albicans (ATCC and field strain), and C. glabrata. The combination method was better than the chemical method for E. coli and C. tropicalis, and the mechanical method showed intermediate results. Conclusion: The three denture hygiene methods showed different effects depending on the type of microbial biofilms formed on acrylic base resin specimens.

An improved conflicting evidence combination approach based on a new supporting probability distance

To avoid counter-intuitive result of classical Dempster's combination rule when dealing with highly conflict information, many improved combination methods have been developed through modifying the basic probability assignments (BPAs) of bodies of evidence (BOEs) by using a certain measure of the degree of conflict or uncertain information, such as Jousselme's distance, the pignistic probability distance and the ambiguity measure. However, if BOEs contain some non-singleton elements and the differences among their BPAs are larger than 0.5, the current conflict measure methods have limitations in describing the interrelationship among the conflict BOEs and may even lead to wrong combination results. In order to solve this problem, a new distance function, which is called supporting probability distance, is proposed to characterize the differences among BOEs. With the new distance, the information of how much a focal element is supported by the other focal elements in BOEs can be given. Also, a new combination rule based on the supporting probability distance is proposed for the combination of the conflicting evidences. The credibility and the discounting factor of each BOE are generated by the supporting probability distance and the weighted BOEs are combined directly using Dempster's rules. Analytical results of numerical examples show that the new distance has a better capability of describing the interrelationships among BOEs, especially for the highly conflicting BOEs containing non-singleton elements and the proposed new combination method has better applicability and effectiveness compared with the existing methods.

Long-term stability of adipose tissue generated from a vascularized pedicled fat flap inside a chamber

Background Numerous studies demonstrate the generation and short-term survival of adipose tissue; however, long-term persistence remains elusive. This study evaluates long-term survival and transferability of de novo adipose constructs based on a ligated vascular pedicle and tissue engineering chamber combination. Methods Defined adipose tissue flaps were implanted into rats in either intact or perforated domed chambers. In half of the groups, the chambers were removed after 10 weeks and the constructs transferred on their vascular pedicle to a new site, where they were observed for a further 10 weeks. In the remaining groups, the tissue construct was observed for 20 weeks inside the chamber. Tissue volume was assessed using magnetic resonance imaging and histologic measures, and constructs were assessed for stability and necrosis. Sections were assessed histologically and for proliferation using Ki-67. Results At 20 weeks, volume analysis revealed an increase in adipose volume from 0.04 ± 0.001 ml at the time of insertion into the chambers to 0.27 ± 0.004 ml in the closed and 0.44 ± 0.014 ml in the perforated chambers. There was an additional increase of approximately 10 to 15 percent in tissue volume in flaps that remained in chambers for 20 weeks, whereas the volume of the transferred tissue not in chambers remained unaltered. Histomorphometric assessment of the tissues documented no signs of hypertrophy, fat necrosis, or atypical changes of the newly generated tissue. Conclusion This study presents a promising new method of generating significant amounts of mature, vascularized, stable, and transferable adipose tissue for permanent autologous soft-tissue replacement.

Structure Discovery in Nonparametric Regression through Compositional Kernel Search

Despite its importance, choosing the structural form of the kernel in nonparametric regression remains a black art. We define a space of kernel structures which are built compositionally by adding and multiplying a small number of base kernels. We present a method for searching over this space of structures which mirrors the scientific discovery process. The learned structures can often decompose functions into interpretable components and enable long-range extrapolation on time-series datasets. Our structure search method outperforms many widely used kernels and kernel combination methods on a variety of prediction tasks.

Algebraic Functions For Recognition

In the general case, a trilinear relationship between three perspective views is shown to exist. The trilinearity result is shown to be of much practical use in visual recognition by alignment --- yielding a direct method that cuts through the computations of camera transformation, scene structure and epipolar geometry. The proof of the central result may be of further interest as it demonstrates certain regularities across homographies of the plane and introduces new view invariants. Experiments on simulated and real image data were conducted, including a comparative analysis with epipolar intersection and the linear combination methods, with results indicating a greater degree of robustness in practice and a higher level of performance in re-projection tasks.

Combining probability forecasts

We consider different methods for combining probability forecasts. In empirical exercises, the data generating process of the forecasts and the event being forecast is not known, and therefore the optimal form of combination will also be unknown. We consider the properties of various combination schemes for a number of plausible data generating processes, and indicate which types of combinations are likely to be useful. We also show that whether forecast encompassing is found to hold between two rival sets of forecasts or not may depend on the type of combination adopted. The relative performances of the different combination methods are illustrated, with an application to predicting recession probabilities using leading indicators.

Utilizando Pesos estáticos e dinâmicos em sistemas multi-classificadores com diferentes níveis de diversidade

Although some individual techniques of supervised Machine Learning (ML), also known as classifiers, or algorithms of classification, to supply solutions that, most of the time, are considered efficient, have experimental results gotten with the use of large sets of pattern and/or that they have a expressive amount of irrelevant data or incomplete characteristic, that show a decrease in the efficiency of the precision of these techniques. In other words, such techniques can t do an recognition of patterns of an efficient form in complex problems. With the intention to get better performance and efficiency of these ML techniques, were thought about the idea to using some types of LM algorithms work jointly, thus origin to the term Multi-Classifier System (MCS). The MCS s presents, as component, different of LM algorithms, called of base classifiers, and realized a combination of results gotten for these algorithms to reach the final result. So that the MCS has a better performance that the base classifiers, the results gotten for each base classifier must present an certain diversity, in other words, a difference between the results gotten for each classifier that compose the system. It can be said that it does not make signification to have MCS s whose base classifiers have identical answers to the sames patterns. Although the MCS s present better results that the individually systems, has always the search to improve the results gotten for this type of system. Aim at this improvement and a better consistency in the results, as well as a larger diversity of the classifiers of a MCS, comes being recently searched methodologies that present as characteristic the use of weights, or confidence values. These weights can describe the importance that certain classifier supplied when associating with each pattern to a determined class. These weights still are used, in associate with the exits of the classifiers, during the process of recognition (use) of the MCS s. Exist different ways of calculating these weights and can be divided in two categories: the static weights and the dynamic weights. The first category of weights is characterizes for not having the modification of its values during the classification process, different it occurs with the second category, where the values suffers modifications during the classification process. In this work an analysis will be made to verify if the use of the weights, statics as much as dynamics, they can increase the perfomance of the MCS s in comparison with the individually systems. Moreover, will be made an analysis in the diversity gotten for the MCS s, for this mode verify if it has some relation between the use of the weights in the MCS s with different levels of diversity

Nausea e vomito em laparoscopia ginecologica: Efeitos do oxido nitroso e da metoclopramida

Background and Objectives - Gynecological laparoscopy causes high postoperative morbidity, mainly due to occurrences such as nausea and vomiting. They result from a great multiplicity of etiologies and drugs used in anesthesia may function as contributing factors. Both the emetic properties of nitrous oxide and the efficacy of metoclopramide as antiemetic agent are controversial. This study was undertaken to determine the effects of both drugs, when used alone or in combination. Methods - Eighty three physical status ASA I and II women were studied. They were premedicated with midazolam before induction of anesthesia with alfentanil and propofol. Anesthesia was maintained with isoflurane with or without nitrous oxide in oxygen. Muscle relaxation was achieved with atracurium. There were 4 groups of patients: GI: midazolam, alfentanil, propofol, atracurium, isoflurane/oxygen; GII: midazolam, alfentanil, propofol, atracurium, isoflurane/nitrous oxide/oxygen; GIII: metoclopramide, midazolam, alfentanil, propofol, atracurium, isoflurane/oxygen; GIV: metoclopramide, midazolam, alfentanil, propofol, atracurium, isoflurane/nitrous oxide/oxygen. The incidence of nausea and vomiting was assessed both in the recovery room (RR) and in the ward. Results - There were no significant differences as regards age, weight and height of the patients and duration of anesthesia and surgery. Nausea and vomiting were more frequent in patients who received N2O (GII, 50%; GIV, 33%), as compared to those who didn't receive this agent (GI and GII, 9.5% and 14.35%, respectively). Metoclopramide decreased the incidence of nausea and vomiting in the recovery room, in patients who didn't receive N2O (GII). These patients remained in the recovery room for 90 minutes. Conclusions - N2O increases the incidence of nausea and vomiting and metoclopramide is effective in reducing these complications only in the recovery room.

Hormone-regulated expression and distribution of versican in mouse uterine tissues

Abstract Background Remodeling of the extracellular matrix is one of the most striking features observed in the uterus during the estrous cycle and after hormone replacement. Versican (VER) is a hyaluronan-binding proteoglycan that undergoes RNA alternative splicing, generating four distinct isoforms. This study analyzed the synthesis and distribution of VER in mouse uterine tissues during the estrous cycle, in ovariectomized (OVX) animals and after 17beta-estradiol (E2) and medroxyprogesterone (MPA) treatments, either alone or in combination. Methods Uteri from mice in all phases of the estrous cycle, and animals subjected to ovariectomy and hormone replacement were collected for immunoperoxidase staining for versican, as well as PCR and quantitative Real Time PCR. Results In diestrus and proestrus, VER was exclusively expressed in the endometrial stroma. In estrus and metaestrus, VER was present in both endometrial stroma and myometrium. In OVX mice, VER immunoreaction was abolished in all uterine tissues. VER expression was restored by E2, MPA and E2+MPA treatments. Real Time PCR analysis showed that VER expression increases considerably in the MPA-treated group. Analysis of mRNA identified isoforms V0, V1 and V3 in the mouse uterus. Conclusion These results show that the expression of versican in uterine tissues is modulated by ovarian steroid hormones, in a tissue-specific manner. VER is induced in the myometrium exclusively by E2, whereas MPA induces VER deposition only in the endometrial stroma.

Randomized placebo-controlled trial of ursodeoxycholic acid with vitamin e in nonalcoholic steatohepatitis

BACKGROUND ; AIMS: Nonalcoholic steatohepatitis (NASH) is a frequent liver disease that can progress to cirrhosis and for which there is no recognized therapy. UDCA and vitamin E have been considered separately as therapeutic options and have not been shown to be effective. This study tested their combination. METHODS: Patients with elevated aminotransferase levels and drinking less than 40 g alcohol/week with biopsy-proven NASH were randomly assigned to receive UDCA 12-15 mg.kg-1.day-1 with vitamin E 400 IU twice a day (UDCA/Vit E), UDCA with placebo (UDCA/P), or placebo/placebo (P/P). After 2 years, they underwent a second liver biopsy. Biopsy specimens were collected, blinded, and scored by a single liver pathologist. RESULTS: Forty eight patients were included, 15 in the UDCA/Vit E group, 18 in the UDCA/P group, and 15 in the P/P group; 8 patients dropped out, none because of side effects. Baseline parameters were not significantly different between the 3 groups. Body mass index remained unchanged during the study. Serum aspartate aminotransferase (AST) and alanine aminotransferase (ALT) levels diminished significantly in the UDCA/Vit E group. Neither the AST nor the ALT levels improved in the P/P group and only the ALT levels in the UDCA/P group. Histologically, the activity index was unchanged at the end of the study in the P/P and UDCA/P groups, but it was significantly better in the UDCA/Vit E group, mostly as a result of regression of steatosis. CONCLUSIONS: Two years of treatment with UDCA in combination with vitamin E improved laboratory values and hepatic steatosis of patients with NASH. Larger trials are warranted.

Assessing eco-environmental performance of agricultural production in OECD countries : combination of soil surface, soil system and farm gate methods of nutrient auditing

Nitrogen balance is increasingly used as an indicator of the environmental performance of agricultural sector in national, international, and global contexts. There are three main methods of accounting the national nitrogen balance: farm gate, soil surface, and soil system. OECD (2008) recently reported the nitrogen and phosphorus balances for member countries for the 1985 - 2004 period using the soil surface method. The farm gate and soil system methods were also used in some international projects. Some studies have provided the comparison among these methods and the conclusion is mixed. The motivation of this present paper was to combine these three methods to provide a more detailed auditing of the nitrogen balance and flows for national agricultural production. In addition, the present paper also provided a new strategy of using reliable international and national data sources to calculate nitrogen balance using the farm gate method. The empirical study focused on the nitrogen balance of OECD countries for the period from 1985 to 2003. The N surplus sent to the total environment of OECD surged dramatically in early 1980s, gradually decreased during 1990s but exhibited an increasing trends in early 2000s. The overall N efficiency however fluctuated without a clear increasing trend. The eco-environmental ranking shows that Australia and Ireland were the worst while Korea and Greece were the best.