Abstract
Portugal is committed towards the EU to an ambitious target of attaining 39% of renewable energy production by 2010, as defined in the RES Directive. The new big hydroelectric investments, together with the mini-hydro power stations that are expected to be ready until 2010 will not be able to increase largely this share of needed renewable energy. The missing renewable energy should be produced exploiting wind energy, meaning that by 2010 a wind power capacity of more than 4.000 MW should to be in operation. At the same time Spain is also increasing wind power integration, such that by 2010 more than 13.000 MW are expected to be installed. The Spanish and the Portuguese electric power systems are facing at the same the challenge of a regional electricity market - the Iberian Electricity Market. This requires that wind power integration must be tackled having in mind technical and commercial concerns.

Abstract
One way of exploring protein unfolding events associated with the development of Amyloid diseases is through the use of multiple Molecular Dynamics Protein Unfolding Simulations. The analysis of the huge amount of data generated in these simulations is not a trivial task. In the present report, we demonstrate the use of Association Rules applied to the analysis of the variation profiles of the Solvent Accessible Surface Area of the 127 amino-acid residues of the protein Transthyretin, along multiple simulations. This allowed us to identify a set of 28 hydrophobic residues forming a hydrophobic cluster that might be essential in the unfolding and folding processes of Transthyretin.

Abstract
This paper presents an adaptation of the peepholing method to regression trees. Peepholing was described as a means to overcome the major computational bottleneck of growing classification trees by Catlett [3]. This method involves two major steps: shortlisting and blinkering. The former has the goal of eliminating some continuous variables from consideration when growing the tree, while the second tries to restrict the range of values of the remaining continuous variables that should be considered when searching for the best cut point split. Both are effective means of overcoming the most costly step of growing tree-based models: sorting the values of the continuous variables before selecting their best split. In this work we describe the adaptations that are necessary to use this method within regression trees. The major adaptations involve developing means to obtain biased estimates of the criterion used to select the best split of these models. We present some preliminary experiments that show the effectiveness of our proposal.

Abstract
The formal specification of agent-based and holonic manufacturing control systems assumes a critical role in order to understand and synthesize those complex systems. This paper presents the formal specification of the coordination models for the ADACOR holonic control system. For this purpose, it is used High-Level Petri Nets to model the behavior of individual ADACOR entities, AUML interaction diagrams to represent the interaction between those entities and mailboxes structures to synchronize the evolution of the Petri net models associated to the holons.

Abstract
This paper is concerned with the problem of predicting relative performance of classification algorithms. It focusses on methods that use results on small samples and discusses the shortcomings of previous approaches. A new variant is proposed that exploits, as some previous approaches, meta-learning. The method requires that experiments be conducted on few samples. The information gathered is used to identify the nearest learning curve for which the sampling procedure was carried out fully. This in turn permits to generate a prediction regards the relative performance of algorithms. Experimental evaluation shows that the method competes well with previous approaches and provides quite good and practical solution to this problem.

Abstract
We tackle the problem of sequence classification using relevant subsequences found in a dataset of protein labelled sequences. A subsequence is relevant if it is frequent and has a minimal length. For each query sequence a vector of features is obtained. The features consist in the number and average length of the relevant subsequences shared with each of the protein families. Classification is performed by combining these features in a Bayes Classifier. The combination of these characteristics results in a multi-class and multi-domain method that is exempt of data transformation and background knowledge. We illustrate the performance of our method using three collections of protein datasets. The performed tests showed that the method has an equivalent performance to state of the art methods in protein classification.