Abstract

Abstract
Classification datasets, which feature a skewed class distribution, are said to be class imbalance. Traditional methods favor the larger classes. We propose pairwise ranking as a method for imbalance classification so that learning compares pairs of observations from each class, and therefore both contribute equally to the decision boundary. In previous work, we suggested treating the binary classification as a ranking problem, followed by a threshold mapping to convert back the ranking score to the original classes. In this work, the method is extended to multi-class ordinal classification, and a new mapping threshold is proposed. Results are compared with traditional and ordinal SVMs, and ranking obtains competitive results.

Abstract
Microscopy examination has been the pillar of malaria diagnosis, being the recommended procedure when its quality can be maintained. However, the need for trained personnel and adequate equipment limits its availability and accessibility in malaria-endemic areas. Rapid, accurate, accessible diagnostic tools are increasingly required, as malaria control programs extend parasite-based diagnosis and the prevalence decreases. This paper presents an image processing and analysis methodology using supervised classification to assess the presence of malaria parasites and determine the species and life cycle stage in Giemsa-stained thin blood smears. The main differentiation factor is the usage of microscopic images exclusively acquired with low cost and accessible tools such as smartphones, a dataset of 566 images manually annotated by an experienced parasilogist being used. Eight different species-stage combinations were considered in this work, with an automatic detection performance ranging from 73.9% to 96.2% in terms of sensitivity and from 92.6% to 99.3% in terms of specificity. These promising results attest to the potential of using this approach as a valid alternative to conventional microscopy examination, with comparable detection performances and acceptable computational times.

Abstract
IntroductionBreast cosmesis is an important endpoint of breast conserving therapy (BCT), but a gold standard method of its evaluation is not yet established. The St. George and Wollongong Randomised Breast Boost trial used five different methods of cosmetic assessment, including both subjective and objective, to comprehensively evaluate the cosmetic outcome of the trial patients. This current study analyses the level of concordance between these methods in an attempt to determine a possible standard in the evaluation of breast cosmesis. MethodsPatients attending follow-up clinic reviews at 5years post breast radiotherapy were evaluated. Patients completed a cosmesis and functional assessment questionnaire, assessing clinicians completed an EORTC (European Organization for Research and Treatment of Cancer) cosmetic rating questionnaire and photographs were obtained. The photographs were later assessed by a panel of five experts, as well as analysed using the objective pBRA (relative Breast Retraction Assessment) and the BCCT.core (Breast Cancer Conservative Treatment.cosmetic results) computer software. Scores were dichotomised to excellent/good and fair/poor. Pairwise comparisons between all methods, except pBRA, were carried out using overall agreement calculations and kappa scores. pBRA scores were compared on a continuous scale with each of the other dichotomised scores obtained by the other four methods. ResultsOf 513 St George patients alive at 5years, 385 (75%) attended St George for follow-up and consented to photography. Results showed that assessment by physicians in clinic and patient self-assessment were more favourable regarding overall cosmetic outcome than evaluation of photographs by the panel or the BCCT.core software. Excellent/good scores by clinician-live and patient self-assessments were 93% and 94% respectively (agreement 89%), as compared to 75% and 74% only by BCCT.core and panel assessments respectively (agreement 83%, kappa 0.57). For the pBRA measurements, there was a statistically significant difference (P <0.001) between scores for excellent/good versus fair/poor cosmesis by all four methods. The range of median pBRA measurements for fair/poor scores was 13.4-14.8 and for excellent/good scores was 8.0-9.4. ConclusionIncorporating both BCCT.core assessment and patient self-assessment could potentially provide the basis of a gold standard method of breast cosmetic evaluation. BCCT.core represents an easy, time efficient, reproducible, cost effective and reliable method, however, it lacks the functional and psychosocial elements of cosmesis that only patient self-reported outcomes can provide.

Abstract
Nowadays imbalanced learning represents one of the most vividly discussed challenges in machine learning. In these scenarios, one or some of the classes in the problem have a significantly lower a priori probability, usually leading to trivial or non-desirable classifiers. Because of this, imbalanced learning has been researched to a great extent by means of different approaches. Recently, the focus has switched from binary classification to other paradigms where imbalanced data also arise, such as ordinal classification. This paper tests the application of learning pairwise ranking with multiple granularity levels in an ordinal and imbalanced classification problem where the aim is to construct an accurate model for donor-recipient allocation in liver transplantation. Our experiments show that approaching the problem as ranking solves the imbalance issue and leads to a competitive performance.

Abstract
Sign Language Recognition (SLR) has becoming one of the most important research areas in the field of human computer interaction. SLR systems are meant to automatically translate sign language into text or speech, in order to reduce the communicational gap between deaf and hearing people. The aim of this paper is to exploit multimodal learning techniques for an accurate SLR, making use of data provided by Kinect and Leap Motion. In this regard, single-modality approaches as well as different multimodal methods, mainly based on convolutional neural networks, are proposed. Experimental results demonstrate that multimodal learning yields an overall improvement in the sign recognition performance.