Abstract
The wine sector requires quick and reliable methods for Vitis vinifera L. varietal identification. The number of V. vinifera varieties is estimated in about 5,000 worldwide. Single Nucleotide Polymorphisms (SNPs) represent the most basic and abundant form of genetic sequence variation, being adequate for varietal discrimination. The aim of this work was to develop DNA-based assays suitable to detect SNP variation in V. vinifera, allowing varietal discrimination. Genotyping by sequencing allowed the detection of eleven SNPs on two genes of the anthocyanin pathway, the flavanone 3-hydroxylase (F3H, EC: 1.14.11.9), and the leucoanthocyanidin dioxygenase (LDOX, EC 1.14.11.19; synonym anthocyanidin synthase, ANS) in twenty V. vinifera varieties. Three High Resolution Melting (HRM) assays were designed based on the sequencing information, discriminating five of the 20 varieties: Alicante Bouschet, Donzelinho Tinto, Merlot, Moscatel Galego and Tinta Roriz. Sanger sequencing of the HRM assay products confirmed the HRM profiles. Three probes, with different lengths and sequences, were used as bio-recognition elements in an optical biosensor platform based on a long period grating (LPG) fiber optic sensor. The label free platform detected a difference of a single SNP using genomic DNA samples. The two different platforms were successfully applied for grapevine varietal identification.

Abstract
The energy demand of the residential sector and the adjacent option for fossil fuels has negative consequences by both greenhouse gases (GHG) and other air pollutants emissions. Since home energy demand consists mainly of energy requirements for space and water heating along with the energy dedicated for appliances, different strategies that aim to stimulate an efficient use of energy need to be reinforced at all levels of human activity. In this paper, a comprehensive comparison is made between the thermostat (ON/OFF), proportional-integral-derivative (PID) and Model Predictive Control (MPC) control models of a domestic heating, ventilation and air conditioning (HVAC) system controlling the temperature of a room. A power interface that adjusts the MPC dynamic range of the output command signal into a discrete two level control signal is proposed, as a new contribution to earlier studies. The model of the house with local solar microgeneration is assumed to be located in a Portuguese city. The household of the case study is subject to the local solar irradiance, temperature and 5 Time-of-Use (Toll) electricity rates applied on an entire week of August 2016. The purpose of the optimisation is to achieve the best compromise between temperature comfort levels and energy costs and also to assess which is the best electricity ToU rate option provided by the electricity retailer for the residential sector. Also, for each electrical load of the HVAC system, the energy and cost are calculated and the results are presented by varying the different MPC weight combination in order to obtain the best possible solution and increase the quality of the model. Finally, after the best tariff and controller are determined, the impact of the solar generation is assessed.

Abstract
Text Mining and NLP techniques are a hot topic nowadays. Researchers thrive to develop new and faster algorithms to cope with larger amounts of data. Particularly, text data analysis has been increasing in interest due to the growth of social networks media. Given this, the development of new algorithms and/or the upgrade of existing ones is now a crucial task to deal with text mining problems under this new scenario. In this paper, we present an update to TextRank, a well-known implementation used to do automatic keyword extraction from text, adapted to deal with streams of text. In addition, we present results for this implementation and compare them with the batch version. Major improvements are lowest computation times for the processing of the same text data, in a streaming environment, both in sliding window and incremental setups. The speedups obtained in the experimental results are significant. Therefore the approach was considered valid and useful to the research community.

Abstract
Two multi-path interferometers were developed using cleaved silica microspheres. A microsphere on top of a singlemode fiber tip was cleaved with a focused ion beam. The asymmetry introduced in the structure generates a new set of optical paths due to random reflections inside the microsphere. The obtained reflection spectrum presents a random-like interferometric behavior with strong spectral modulation of around 3 dB amplitude. Two distinct regions can be observed when a fast Fourier transform is applied. The first involves two cavities at a lower frequency and the second region involves a band of frequencies that is originated by the random interferometric reflections. These two spectral characteristics can be separated using low-pass and high-pass filters, respectively. A correlation method was used to obtain a temperature response from the two-cavity component. A similar structure was also created in a microsphere of multimode fiber. The microsphere was cleaved by polishing the structure with a certain angle. The interference between the different optical paths can be seen as the superposition of several two-wave interferometers, which can be discriminated through signal processing. Temperature sensing was also explored with this structure. The sensitivity to temperature is more than 3-fold for smaller cavities. Moreover, a sensitivity enhancement is also verified if a correlation method is used.

Abstract
Pose invariant object detection and classification plays a critical role in robust image recognition systems and can be applied in a multitude of applications, ranging from simple monitoring to advanced tracking. This paper analyzes the usage of the Bag of Words model for recognizing objects in different scales, orientations and perspective views within cluttered environments. The recognition system relies on image analysis techniques, such as feature detection, description and clustering along with machine learning classifiers. For pinpointing the location of the target object, it is proposed a multiscale sliding window approach followed by a dynamic thresholding segmentation. The recognition system was tested with several configurations of feature detectors, descriptors and classifiers and achieved an accuracy of 87% when recognizing cars from an annotated dataset with 177 training images and 177 testing images.

Abstract