Abstract
Accounting for misclassification costs is important in many practical applications of machine learning, and cost sensitive techniques for classification have been studied extensively. Utility-based learning provides a generalization of purely cost-based approaches that considers both costs and benefits, enabling application to domains with complex cost-benefit settings. However, there is little work on utility- or cost-based learning for regression. In this paper, we formally define the problem of utility-based regression and propose a strategy for maximizing the utility of regression models. We verify our findings in a large set of experiments that show the advantage of our proposal in a diverse set of domains, learning algorithms and cost/benefit settings.

Abstract

Abstract
Entrepreneurs are critical for countries' industrial dynamics and economic growth. Although there are several studies focusing on the attitudes, cognitions and traits of entrepreneurs, only a few have explored the link between entrepreneurship and the job features entrepreneurs most value in the work setting. Taking a longitudinal perspective, over two periods of time, 1990-1993 and 2008-2010, the present study resorted to the European Value Study dataset, encompassing 27 European countries and more than 100 thousand individuals. Based on logistic estimates, we found that, compared with non-entrepreneurs, entrepreneurs consider work more important than other life dimensions (except family). Moreover, some work values associated with entrepreneurship (e.g., use initiative, achieving something, responsible job, meeting abilities, have a say, and learning new skills) are indeed mentioned the most by entrepreneurs and emerged significantly related with entrepreneurial propensity. Finally, some work values (e.g., have a say) are associated with entrepreneurship propensity regardless of country's culture, whereas others (e.g., interesting job) are strongly influenced by culture.

Abstract
Monitoring and managing performance in the workplace is nowadays an important aspect, in a time in which methodologies like Agile push individual and team limits further. Current performance monitoring approaches are either intrusive or based on productivity measures and are thus often dreaded by workers. Moreover, these approaches do not take into account the importance and role of the numerous external factors that influence productivity. We present a non-intrusive performance monitoring environment based on behavioral biometrics and real time analytics. It monitors and analyzes 15 features extracted from the workers' interaction with the computer and can provide a measure of performance that is completely transparent. This measure is sensitive to external factors such as mental fatigue, stress or emotional valence. We validate this environment by assessing the effects of musical selection on Human-Computer Interaction. Results show a significant improvement on mouse motion when participants listen to the selected auditory stimuli and a negative effect on typing performance, especially with stimuli with positive tension. This work will enable the development of performance monitoring and management environments, with benefits for both organizations and individuals.

Abstract
This paper presents a new methodology to define a priority scale for maintenance actions in substations, based on the development of a Composite Risk Index (CRI) associated with each device. Two auxiliary indices are built: Basic Condition (BC) and Operating Condition (OC), representing the physical and functional characteristics of the equipment that can compromise their performance and contribute to the occurrence of failures. Their evaluation is helped by a Technical Capacity Index (TCI), which evaluates how much the equipment has been affected by wear and tear, in the assessment of the Basic Condition, and the classification of the equipment defects by degrees of severity, in the assessment of the Operating Condition. Two cascading Fuzzy Inference Systems of the Mandani type are used, the first in defining the BC, and the second to obtain the equipment CRI denoting maintenance priority, which may then be used in planning maintenance actions. The methodology is verified through an SF6 circuit breaker CRI assessment, and its priority scale for maintenance planning. The method for evaluating the SF6 circuit breakers reliability is validated through a comparison with a statistical approach, using real data collected from equipment installed in Eletrobras Eletronorte Transmission System, in Rondonia, Amazon region of Brazil. (C) 2016 Published by Elsevier B.V.

Abstract
In this paper a modified Newton type algorithm (NTA) is analysed to estimate the electrical signal spectral composition. As core of this study, it is given attention to the accuracy and sensitivity of the algorithm for base frequency and harmonic content characterization under noisy measurements. Convergence speed, stability issues and computational complexity of the implementation are detailed and compared using clean and noisy test signals. Conclusions are duly drawn.