Abstract
This paper introduces a novel strategy for the task of simultaneously locating two key anatomical landmarks in retinal images of the eye fundus, namely the optic disc and the fovea. For that, instead of attempting to classify each pixel as belonging to the background, the optic disc, or the fovea center, which would lead to a highly class-imbalanced setting, the problem is reformulated as a pixelwise regression task. The regressed quantity consists of the distance from the closest landmark of interest. A Fully-Convolutional Deep Neural Network is optimized to predict this distance for each image location, implicitly casting the problem into a per-pixel Multi-Task Learning approach by which a globally consistent distribution of distances across the entire image can be learned. Once trained, the two minimal distances predicted by the model are selected as the locations of the optic disc and the fovea. The joint learning of every pixel position relative to the optic disc and the fovea favors an automatic understanding of the overall anatomical distribution. This results in an effective technique that can detect both locations simultaneously, as opposed to previous methods that handle both tasks separately. Comprehensive experimental results on a large public dataset validate the proposed approach.

Abstract
In active distribution grids (ADNs), a distribution corporation (Disco) can trade electricity with micro-grids (MGs) besides trading electricity with wholesale markets. Therefore, the operational flexibility of the Disco is increased so that it can play as a price-maker agent in electricity markets. To model the actions of Disco, a bi-level optimization method is developed where Disco problem is modeled at the upper-level problem, whereas the MGs problem together with the day-ahead market clearing procedure are modeled as the lower-level problems. To take the stochastic performance of green energy integration and loads into account, the Disco analysis is shown as two-stage stochastic problem, in which the Disco risk aversion is programmed considering the conditional value-at-risk tool. The subsequent non-linear bi-level approach is converted into a linear single-level approach through Karush-Kuhn-Tucker (KKT) conditions and dual theory. To validate the success of the proposed method, a 24-bus power system is used. Conclusions are duly drawn.

Abstract
The European Union (EU) is adopting proactive strategies toward global decarbonization, proposing ambitious climate objectives to the international community, and adopting by itself ambitious energy and climate change objectives, as can be checked in its strategic packages for 2020, 2030 and 2050. As can be seen in these packages and in global climate and energy agreements, renewable generation is a key aspect to reach the global decarbonizing objectives. This chapter starts with a review of the very basic concepts of greenhouse gases emissions and its impact on climate change and summarizes the main objectives of the strategic EU energy packages and of the Paris agreement. It then focuses on methodologies for estimating the impact of renewable energies on greenhouse gases emissions reduction and reviews the reported EU related achievements. It also proposes a classification of EU countries in terms of Tapio decoupling states by analyzing and classifying the countries emissions intensities and its variation for a considered time period and analyzes the impact renewable energies had in that greenhouse gases emissions variations, in relation with other possibly relevant factors. The chapter ends with a summary of the expected evolution of renewable energies in the EU, and with the final conclusions.

Abstract

Abstract

Abstract
Technology is constantly evolving, as a result, users have become more demanding and the applications more complex. In the realm of Web development, JavaScript is growing in a surprising way, already leaving the boundaries of the browser, mainly due to the advent of Node.js. In fact, JavaScript is constantly being reinvented and, from the ES2015 version, began to include the OO concepts typically found in other programming languages. With Web access being mostly made by mobile devices, developers face now performance challenges and need to perform a plethora of tasks that weren’t necessary a decade ago, such as managing dependencies, bundling files, minifying code, optimizing images and others. Many of these tasks can be achieved by using the right tools for the job. However, developers not only have to know those tools, but they also must know how to access and operate them. This process can be tedious, confusing, time-consuming and error-prone. In this paper, we present Kaang, an automatic generator of RESTFul Web applications. The ultimate goal of Kaang is to minimize the impact of creating a RESTFul service by automating all its workflow (e.g., files structuring, boilerplate code generation, dependencies management, and task building). This kind of generators will benefit two types of users: will help novice developers to decrease their learning curve while facing the new frameworks and libraries commonly found in the modern Web and speed up the work of expert developers avoiding all the repetitive and bureaucratic work. At the same time, Kaang promotes the good development principles by adding automatic testing and documentation generation. For this accomplishment, Kaang generates the main API content based on the user’s input and a set of templates which will help developers to manage and test routes, define resources, store data models and others. In order to provide an addition level of confidence to the generator’s end-users, the generator will be integrated on Travis CI and published on both the npmjs and Yeoman registries. © Ricardo Queirós.