Abstract

Abstract
Overhead Power Distribution Lines (OPDLs) correspond to a large percentage of the medium-voltage electrical systems. In these networks, visual inspection activities are usually performed without resorting to automated systems, requiring a significant investment of time and human resources. We present a methodology to identify the defect and type of insulators using Convolutional Neural Networks (CNNs). More than 2500 photographs were collected both from inside a studio and from a realistic OPDL. A classification model is proposed to automatically recognize the insulators conformity. This model is able to learn from indoors photographs by augmenting these images with realistic details such as top ties and real-world backgrounds. Furthermore, Multi-Task Learning (MTL) was used to improve performance of defect detection by also predicting the insulator class. The proposed methodology is able to achieve an accuracy of 92% for material classification and 85% for defect detection, with F1-score of 0.75, surpassing available solutions.

Abstract

Abstract
The majority of computer-Aided diagnosis methods for Alzheimer's disease (AD) from brain images either address only two stages of the disease at a time (and reduce the problem to binary classification) or do not exploit the ordinal nature of the different classes. An exception is the work by Fan et al. [1], which proposed an ordinal method that obtained better performance than traditional multiclass classification. Still, special care should be taken when data is class imbalanced, i.e. when some classes are overly represented when compared to others. Building on top of [1], this work makes use of a recently published ordinal classifier, which transforms the problem into sets of pairwise ranking problems, in order to address the class imbalance in the data [2]. Several methods were experimented with, using a Support Vector Machine as the underlying estimator. The pairwise ranking approach has shown promising results, both for traditional and imbalance metrics. © 2018 IEEE.

Abstract
Imbalanced classification has been extensively researched in the last years due to its prevalence in real-world datasets, ranging from very different topics such as health care or fraud detection. This literature has long been dominated by variations of the same family of solutions (e.g. mainly resampling and cost-sensitive learning). Recently, a new and promising way of tackling this problem has been introduced: learning with scoring pairwise ranking so that each pair of classes contribute in tandem to the decision boundary. In this sense, the paper addresses the problem of class imbalance in the context of ordinal regression, proposing two novel contributions: (a) approaching the imbalance by binary pairwise ranking using a well-known label decomposition ensemble, and (b) introducing a regularization into this ensemble so that parallel decision boundaries are favored. These are two independent contributions that synergize well. Our model is tested using linear Support Vector Machines and our results are compared against state-of-the-art models. Both approaches show promising performance in ordinal class imbalance, with an overall 15% improvement relative to the state-of-the-art, as evaluated by a balanced metric. © 2018 Springer-Verlag London Ltd., part of Springer Nature