Abstract
Although cross-validation is a standard procedure for performance evaluation, its joint application with oversampling remains an open question for researchers farther from the imbalanced data topic. A frequent experimental flaw is the application of oversampling algorithms to the entire dataset, resulting in biased models and overly-optimistic estimates. We emphasize and distinguish overoptimism from overfitting, showing that the former is associated with the cross-validation procedure, while the latter is influenced by the chosen oversampling algorithm. Furthermore, we perform a thorough empirical comparison of well-established oversampling algorithms, supported by a data complexity analysis. The best oversampling techniques seem to possess three key characteristics: use of cleaning procedures, cluster-based example synthetization and adaptive weighting of minority examples, where Synthetic Minority Oversampling Technique coupled with Tomek Links and Majority Weighted Minority Oversampling Technique stand out, being capable of increasing the discriminative power of data.

Abstract
Half-cycle Posicast Control is currently used in a vast range of applications. Although the proved benefits of this technique, one of its major disadvantages concerns model uncertainties. This has motivated the development and integration of robust methods to overcome this issue. In this paper, a practical experiment for auto-tuning of a two degrees of freedom control configuration using a Half-Cycle Posicast pre-filter (or input-shaping), and a PID controller under parametric variations is presented. The proposed method requires using an oscillatory system model in an auto-tuning control structure. The error derivative among the model and system output is used to trigger both the identification and retuning procedure. The proposed method is flexible for choosing identification plus optimization methods. Practical results obtained for electronic filter plants suggest improved performance for the considered cases. © 2018

Abstract
The use of electronic health records (EHR) to support clinical practices is widespread worldwide, due to the need to optimize health care delivery. Therefore, the usability assessment of EHR systems is crucial. The objective of this study was to perform a qualitative and quantitative assessment of the usability of SClinico, the most used EHR system within the Portuguese National Health Service. This observational study to assess SClinico usability took place in several clinical services of the Centro Hospitalar de Trás-os-Montes e Alto Douro. The results show that SClinico has some usability issues that influence the clinical practice and, therefore, need to be improved.

Abstract
BackgroundThe structure, function, and capacity of the health care system in the Occupied Palestinian Territories (OPT) had been largely shaped by the complex political history of the country. Since the establishment of the Palestinian Authority in 1994, the reform efforts were subsidized much by the international aids to rebuild the country's institutional capacity. No previous studies have provided a realistic evaluation of Palestinian achievements in the conduct of public healthcare, we examine the relative productive efficiency of public hospitals (their managers' success in the usage of resources) during 2010-2015 within West Bank and Jordan. Then, we estimate the efficiency of policies within which managers operate (the program efficiency) across the two countries.MethodsWe employ the Data Envelopment Analysis (DEA) models to distinguish between within-country managerial efficiencies and public policy program efficiencies across the two countries. The study follows two key steps, the first step evaluates managerial efficiencies and explores trends in performance within each country. Then, we examine the program efficiencies across the two countries.ResultsPublic hospitals improved their year-specific overall efficiency from 75 to 80% in the West Bank and from 78 to 86% in Jordan in 2010 and 2015 respectively. Changes in efficiency are driven by scale effects in West Bank and by managerial enhancements in Jordan. Program efficiency in West Bank outperformed Jordan during 2010-2012, there was no significant difference in mean program efficiencies between the two countries during 2013-2015.ConclusionsThis work addresses a gap in the DEA literature by empirically investigating the efficiency of public hospitals as distinct from program efficiency in a developing country, namely, Palestine. Findings stimulate hospital managers to enhance potential improvements, policymakers to allocate resources, and international donors to focus on the right adoption of new technology to get better benefits from their considerable investments in public hospitals.

Abstract
This paper addresses the stochastic security constrained unit commitment (SSCUC) problem with flexibility resources for managing the uncertainty of wind power generation (WPG). Departing from the traditional flexibility resources such as the thermal units with fast up/down spinning reserves and transmission switching (TS), this paper explores also the use of demand response (DR) and energy storage (ES) systems in an innovative integrated scheme. The proposed scheme utilizes a stochastic optimization framework to coordinate the flexibility resources dealing with the uncertainty of WPGs and equipment failures. The stochastic optimization model is formulated as a mixed-integer linear programming (MIP), and this problem is large and computationally complex even for medium sized systems. Accordingly, we present a novel accelerating decomposition technique aimed at solving this problem and reducing the number of iterations and CPU time. Numerical simulation results on the modified 6-bus system and on large-scale power systems, i.e. IEEE 118 and 300-bus systems, clearly demonstrate the benefits of applying flexibility resources for uncertainty management and the efficacy of the proposed solution strategy for large-scale systems.

Abstract
One of the main difficulties in the use of deep learning strategies in medical contexts is the training set size. While these methods need large annotated training sets, these datasets are costly to obtain in medical contexts and suffer from intra and inter-subject variability. In the present work, two new pre-processing techniques are introduced to improve a deep classifier performance. First, data augmentation based on co-registration is suggested. Then, multi-scale enhancement based on Difference of Gaussians is proposed. Results are accessed in a public mammogram database, the InBreast, in the context of an ordinal problem, the BI-RADS classification. Moreover, a pre-trained Convolutional Neural Network with the AlexNet architecture was used as a base classifier. The multi-class classification experiments show that the proposed pipeline with the Difference of Gaussians and the data augmentation technique outperforms using the original dataset only and using the original dataset augmented by mirroring the images.