Abstract
Algorithm selection is an important task in different domains of knowledge. Meta-learning treats this task by adopting a supervised learning strategy. Training examples in meta-learning (called meta examples) are generated from experiments performed with a pool of candidate algorithms in a number of problems, usually collected from data repositories or synthetically generated. A meta-learner is then applied to acquire knowledge relating features of the problems and the best algorithms in terms of performance. In this paper, we address an important aspect in meta-learning which is to produce a significant number of relevant meta-examples. Generating a high quality set of meta-examples can be difficult due to the low availability of real datasets in some domains and the high computational cost of labelling the meta-examples. In the current work, we focus on the generation of meta-examples for meta-learning by combining: (1) a promising approach to generate new datasets (called datasetoids) by manipulating existing ones; and (2) active learning methods to select the most relevant datasets previously generated. The datasetoids approach is adopted to augment the number of useful problem instances for meta-example construction. However not all generated problems are equally relevant. Active meta-learning then arises to select only the most informative instances to be labelled. Experiments were performed in different scenarios, algorithms for meta-learning and strategies to select datasets. Our experiments revealed that it is possible to reduce the computational cost of generating meta-examples, while maintaining a good meta-learning performance.

Abstract

Abstract
This paper presents novel intraday session models for price forecasts (ISMPF models) for hourly price forecasting in the six intraday sessions of the Iberian electricity market (MIBEL) and the analysis of mean absolute percentage errors (MAPEs) obtained with suitable combinations of their input variables in order to find the best ISMPF models. Comparisons of errors from different ISMPF models identified the most important variables for forecasting purposes. Similar analyses were applied to determine the best daily session models for price forecasts (DSMPF models) for the day- ahead price forecasting in the daily session of the MIBEL, considering as input variables extensive hourly time series records of recent prices, power demands and power generations in the previous day, forecasts of demand, wind power generation and weather for the day- ahead, and chronological variables. ISMPF models include the input variables of DSMPF models as well as the daily session prices and prices of preceding intraday sessions. The best ISMPF models achieved lower MAPEs for most of the intraday sessions compared to the error of the best DSMPF model; furthermore, such DSMPF error was very close to the lowest limit error for the daily session. The best ISMPF models can be useful for MIBEL agents of the electricity intraday market and the electric energy industry.

Abstract
With the liberalization of the electricity markets, price forecasting has become crucial for the decision-making process of market agents. The unique features of electricity price, such as non-stationary, non-linearity and high volatility make this a very difficult task. For this reason, rather than a simple point forecast, market participants are more interested in a probabilistic forecast that is essential to estimate the uncertainty involved in the price. By focusing on this issue, the aim of this paper is to analyze the impact of external factors in the electricity price and present a methodology for probabilistic forecasting of day-ahead electricity prices from the Iberian electricity market. The models are built using regression techniques and aim to obtain, for each hour, the quantiles of 5% to 95% by steps of 5%.

Abstract
Electricity markets are becoming more competitive, to some extent due to the increasing number of players that have moved from other sectors to the power industry. This is essentially resulting from incentives provided to distributed generation. Relevant changes in this domain are still occurring, such as the extension of national and regional markets to continental scales. Decision support tools have thereby become essential to help electricity market players in their negotiation process. This paper presents a metalearner to support electricity market players in bidding definition. The proposed metalearner uses a dynamic artificial neural network to create its own output, taking advantage on several learning algorithms already implemented in ALBidS (Adaptive Learning strategic Bidding System). The proposed metalearner considers different weights for each strategy, based on their individual performance. The metalearner's performance is analysed in scenarios based on real electricity markets data using MASCEM (Multi-Agent Simulator for Competitive Electricity Markets). Results show that the proposed metalearner is able to provide higher profits to market players when compared to other current methodologies and that results improve over time, as consequence of its learning process.

Abstract
Underwater imaging is being increasingly helpful for the autonomous robots to reconstruct and map the marine environments which is fundamental for searching for pipelines or wreckages in depth waters. In this context, the accuracy of the information obtained from the environment is of extremely importance. This work presents a study about the accuracy of a reconfigurable stereo vision system while determining a dense disparity estimation for underwater imaging. The idea is to explore the advantage of this kind of system for underwater autonomous vehicles (AUV) since varying parameters like the baseline and the pose of the cameras make possible to extract accurate 3D information at different distances between the AUV and the scene. Therefore, the impact of these parameters is analyzed using a metric error of the point cloud acquired by a stereoscopic system. Furthermore, results obtained directly from an underwater environment proved that a reconfigurable stereo system can have some advantages for autonomous vehicles since, in some trials, the error was reduced by 0.05m for distances between 1.125 and 2.675 m.