Abstract

Abstract
This paper presents a Support Vector Machine (SVM) based approach for energy consumption forecasting. The proposed approach includes the combination of both the historic log of past consumption data and the history of contextual information. By combining variables that influence the electrical energy consumption, such as the temperature, luminosity, seasonality, with the log of consumption data, it is possible for the proposed method by find patterns and correlations between the different sources of data and therefore improves the forecasting performance. A case study based on real data from a pilot microgrid located at the GECAD campus in the Polytechnic of Porto is presented. Data from the pilot buildings are used, and the results are compared to those achieved by several states of the art forecasting approaches. Results show that the proposed method can reach lower forecasting errors than the other considered methods. © 2018 IEEE.

Abstract
A sensor based on 2 hollow core microspheres is proposed. Each microsphere was produced separately through fusion splicing and then joined. The resultant structure is a Fabry-Perot interferometer with multiple interferences that can be approximated to a 4-wave interferometer. Strain characterization was attained for a maximum of 1350 mu epsilon, achieving a linear response with a sensitivity of 3.39 +/- 0.04 pm/mu epsilon. The fabrication technique, fast and with no chemical hazards, as opposed to other fabrication techniques, makes the proposed sensor a compelling solution for strain measurements in hash environments.

Abstract
Introduction. Online health forums help to surface and organize patients' knowledge and make it useful for many. They are used by many to seek for advice or to share what they know about health subjects. Being an important communication medium, it's important to understand why and how it is used. Method. In this work we examine and categorize messages of an online health forum, with the purpose of providing a classification scheme that can be used by the research community in future analyses. The definition of the classification scheme was iterative and its inter-rater reliability was assessed twice using Cohen's Kappa statistic. Analysis. The classification scheme arose from a content analysis of 3,399 messages from several communities of an online health forum. Findings. The scheme is divided into four sections of categories and each section has several subcategories, in total there are 23 subcategories. The inter-rater agreement assessment of the scheme showed a good consistency between coders. The majority of the categories has a Cohen's Kappa agreement above 0.4. Conclusion. The proposed classification scheme facilitates the analysis of messages exchanged in online health forums for several purposes, including studies of information seeking.

Abstract
There are still open questions about the effectiveness of Virtual Reality (VR) in Education when compared to conventional learning methods. This paper studies the feasibility of a VR-based learning tool and the possible differences in knowledge retention across a VR learning method and a conventional audio method, when it comes to learning a foreign language. Also, the students’ sense of presence and satisfaction were studied. For such purpose, a user study was conducted and results revealed that while presence and satisfaction were higher in Virtual Reality, the knowledge retention score remains the same across both experimental conditions.

Abstract
The Classifier kNN is largely used in Human Activity Recognition systems. Research efforts have proposed methods to decrease the high computational costs of the original kNN by focusing, e.g., on approximate kNN solutions such as the ones relying on Locality-sensitive Hashing (LSH). However, embedded kNN implementations need to address the target device memory constraints and power/energy consumption savings. One of the important aspects is the constraint regarding the maximum number of instances stored in the kNN learning process (being it offline or online and incremental). This paper presents simple, energy/computationally efficient and real-time feasible schemes to maintain a maximum number of learning instances stored by kNN. Experiments in the context of HAR show the efficiency of our best approaches, and their capability to avoid the kNN storage runs out of training instances for a given activity, a situation not prevented by typical default schemes.